February 1998

806 messages

« January 1998March 1998 »

Messages

Subject: Re: (usr-tc) OSPF
From: MegaZone <megazone@livingston.com>
Date: 1998-02-01 02:22:37
Once upon a time Jeff Mcadams shaped the electrons to say... >Wha? Did I miss this, or was it posted to another list? In any case, It was on portmster-users, Garry is there too and I know he saw the thread. >good luck at GTE/BBN. Thanks. -MZ
Subject: Re: (usr-tc) OSPF
From: MegaZone <megazone@livingston.com>
Date: 1998-02-01 02:29:21
Once upon a time MegaZone shaped the electrons to say... Oops - forget about reply-to, one of the few lists I'm on that uses it... -MZ
Subject: (usr-tc) Analog MLPPP
From: James Wilson <james@cruxnet.com>
Date: 1998-02-01 03:09:38
Anybody have any luck with analog mulitlink PPP. Is this even supported in the TC hub? I've seen reports of it being used in Ascend boxes. I tried it with two Sportsters and right after authentication, when the MLPPP negotiation starts, my computer reboots. This is with DUN 1.2. I have a client that is really interested in this. James
Subject: Re: (usr-tc) HyperARC/Authentication
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-01 05:24:51
On Sun, 1 Feb 1998, Brian wrote: > On Sun, 1 Feb 1998, jcondon wrote: > > > Brian wrote: > > > > > Are most of you running PAP only, and is it because you found that > > > CHAP/PAP (EITHER) doesnt work nicely? > > > > Same thing on the Netserver I modem Plus and apperently USR knows about > > it because when we called with some radius questions the first thing > > they did was have us set it for PAP instead of either. Why this needed > > I am not sure but, we tried either after we got it working and it > > doesn't work that way. Why either is an option when it doesn't apper to > > be possible to make it work makes little sense to me. Maybe it would > > work with local users but, not radius? Or only with USR radius? > > Might shave a NAK or two off negotiation if it did work, since I beleive > most stacks negotiate CHAP before PAP. > > In any case, they shouldnt have "EITHER" as the default if it doesnt work > properly. > Here is the difference: NETServer - By default the NETServer does PAP first and CHAP next. There is a switch you can setup on the NETServer called chapfst if you set chapfst on then the NETServer will do chap first and pap next. Now most of you have Radius users, and most of you use Unix / NT based radius server who use system password file for password. In these case you cannot do chap, for chap the users should be on your radius users file with passwords. HiPer ARC - It has three options, PAP CHAP EITHER, and also remember HiPer ARC will do MD4 ( microsoft chap also ). Now HiPer ARC will work in the following way. CHAP - Now when a users dials in it will do only chap, if user cannot do chap it will fail. Either - When the user dials in he has to option for starting pap or chap if he does not then HiPer ARC will start chap PAP - The user again has the option to start pap or chap if they do not HiPer arc will start pap. That is the difference, by setting it to either it does do PAP first it just requires that the client is capable of starting either pap or chap. When you ask support they would say set it to pap since if the user does not ask Hiper ARC will do pap first but if the user say no to pap and starts chap HiPer arc will cooperate. krish > Brian > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Systems Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: (usr-tc) IP masquerading
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-01 06:18:50
Hi, I have a question regarding cisco 1000 configuration in this particular environment. I have a lan connected to my Cisco 1000. I use my Cisco 1000 to dial and connect to my ISP. My machines at home connected to my local lan and Cisco 1000, cannot connect to the Internet. What I am looking for with the Cisco 1000 is IP masquerading so that it will allow me to dial into any ISP I want and still be able to use my PC's to connect to internet. Any ideas?? thanks krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) Using quad modems with serial ports
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-01 07:50:54
On Sun, 1 Feb 1998, Brian Elfert wrote: > My PRI NAC has T1 code loaded on it, and is set to autobusy channels. > > I want to hook some of my quad modems to a Portmaster through serial ports > on a NIC card. > > The problem is, after I deactivate the modem in the Netserver and reset > the port, the channels on that card go into busyout status. I got one out > of busyout, and working with the serial port, but no idea how. > > Any suggestions to get this working? Well first you must choose the modem which you want to attach to the Port master, It is not advisable to split a quad modem - what I mean is that if you have a Quad you may want to use all the four modems on the quad to the port master or to the NETServer, if you split 2 on the Port master and 2 on the NETServer you will have to make sure of configuring those properlly. If you use a entier quad then it is easy to setup. Say you want the first four modems on the NETServer to connect to the PM - From the NETServer set those modem inactive and reset them. Set s5-s8 inactive reset s5-s8. Save all Now make sure that the Quad modem is attached to the PM with the octopus cable. Now pull out the Quad card and reset the modem. Now the modem should say its line interface option is TDM but its DTE interface will say is RS232. Now all you have to do is set the modem active on the PM and reset them. krish > > Brian > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) Analog MLPPP
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-01 08:28:57
On Sun, 1 Feb 1998, James Wilson wrote: > Anybody have any luck with analog mulitlink PPP. Is this even supported in > the TC hub? I've seen reports of it being used in Ascend boxes. I tried it > with two Sportsters and right after authentication, when the MLPPP > negotiation starts, my computer reboots. This is with DUN 1.2. I have a > client that is really interested in this. Yes, lots o' luck here. We have multiple customers using DUN 1.2 w/ two modems to establish MLPPP. Our official support policy on analog MLPPP is "hands-off". In other words, "Yes, we know it will work, but it's up to you to figure out how to make it work." Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) list of users via SNMP? No.
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-01 09:09:30
: Is it possible to get a list of users that are login via SNMP? Not with the SNMP agents that are included with TC equipment. The reigning solution to automated retrieval of this information is tcwho and/or pmwho, which (as I understand them) telnet to the netserver card, login as !root, and do a "show sessions". It'd probably be worth the effort to some people to write a proxy agent that'd provide an SNMP interface to the tcwho output. : I didn't find any obvious answers glancing through the SNMP files. Should : this be obtained through NMC or the Netserver? It should be obtained through the Netserver, but Tatai SV Krishnan from USR/3Com says that such a fix would require some rather extensive rewriting of the Netserver code, and they're not likely to do that. I'd be interested to know if the HiPer ARC provides that information via SNMP. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: (usr-tc) HyperARC/Authentication
From: Brian <signal@shreve.net>
Date: 1998-02-01 12:32:15
I noticed that when the HyperARC is set for "EITHER" as its receive authentication value for PPP, that users authentication fails. Yet if i set it for "PAP" it works fine. Why wouldn't either work just as well as PAP? I thought it would try CHAP first and then try PAP and suceed, but users can't login with it set for the default of EITHER. If my understanding is right, I thought the Win95 TCP/IP stack first tried CHAP. I was hoping that by accepting "EITHER" this would speed up connects since there would be one less NAK involved. Are most of you running PAP only, and is it because you found that CHAP/PAP (EITHER) doesnt work nicely? Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Analog MLPPP
From: Brian <signal@shreve.net>
Date: 1998-02-01 12:35:11
On Sun, 1 Feb 1998, System Administrator wrote: > On Sun, 1 Feb 1998, James Wilson wrote: > > > Anybody have any luck with analog mulitlink PPP. Is this even supported in > > the TC hub? I've seen reports of it being used in Ascend boxes. I tried it > > with two Sportsters and right after authentication, when the MLPPP > > negotiation starts, my computer reboots. This is with DUN 1.2. I have a > > client that is really interested in this. > > Yes, lots o' luck here. We have multiple customers using DUN 1.2 w/ two > modems to establish MLPPP. Our official support policy on analog MLPPP is > "hands-off". In other words, "Yes, we know it will work, but it's up to > you to figure out how to make it work." But you do go in there and increase the port limit for that user to 2 channels right? Or does your default RADIUS entry allow for 2 ports to begin with? I like the "hands-off" policy. We may do this, but I would still have to "set" the user up for 2 ports in RADIUS just as I would any MPP connection, and let them take it from there. Brian > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) HyperARC/Authentication
From: jcondon <jcondon@computer.net>
Date: 1998-02-01 13:51:00
Brian wrote: > Are most of you running PAP only, and is it because you found that > CHAP/PAP (EITHER) doesnt work nicely? Same thing on the Netserver I modem Plus and apperently USR knows about it because when we called with some radius questions the first thing they did was have us set it for PAP instead of either. Why this needed I am not sure but, we tried either after we got it working and it doesn't work that way. Why either is an option when it doesn't apper to be possible to make it work makes little sense to me. Maybe it would work with local users but, not radius? Or only with USR radius?
Subject: (usr-tc) Ping won't work
From: Wesley Shih <hshih@smartt.com>
Date: 1998-02-01 14:14:01
Hi, I try to connect one of my ISP, after connected, I issue "ping" command, then my system just hang up there and no response any error message back. I need to do a Crtl-C to break it. So I change phone number, passowrd and username ONLY to dial another ISP, this time everything work great! I try to use win95 to dial the first ISP and without any problem. I am using kernel 3.0.33 ,pppd 2.3.3 to dial up, anf the first ISP told me, they are using 3COM totoal Comtrol manager as terminal server and did NOT turn off the ICMP function in their TC. Anyone can give a hint, what is going on here ? Thank you! Wesley Shih hshih@smartt.com
Subject: RE: (usr-tc) Analog MLPPP
From: Marshall Morgan <marshall@netdoor.com>
Date: 1998-02-01 14:22:03
On Sunday, February 01, 1998 12:35 PM, Brian [SMTP:signal@shreve.net] wrote: > On Sun, 1 Feb 1998, System Administrator wrote: > > > On Sun, 1 Feb 1998, James Wilson wrote: > > > > > Anybody have any luck with analog mulitlink PPP. Is this even supported > > > in > > > the TC hub? I've seen reports of it being used in Ascend boxes. I tried > > > it > > > with two Sportsters and right after authentication, when the MLPPP > > > negotiation starts, my computer reboots. This is with DUN 1.2. I have a > > > client that is really interested in this. > > > > Yes, lots o' luck here. We have multiple customers using DUN 1.2 w/ two > > modems to establish MLPPP. Our official support policy on analog MLPPP is > > "hands-off". In other words, "Yes, we know it will work, but it's up to > > you to figure out how to make it work." > > But you do go in there and increase the port limit for that user to 2 > channels right? Or does your default RADIUS entry allow for 2 ports to > begin with? I assume customers are charged more for MLPPP right? Marshall Morgan http://www.netdoor.com 601.969.1434 Ext. #28 | Fax 601.969.3838 | 800.952.1570 Ext. #28
Subject: (usr-tc) NETServer, routing, proxy-arp, subnet-0
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-01 16:23:11
Hi All, After a bit of advice here... Currently I have NETServer cards at 3.6.28 and 3.5.33, the ptpX and net0 interfaces sit on the same network, and we use proxy-arp. I'm wanting to move the ptpXX interfaces to another network and route the traffic there - I know that bit works. In addition, I need to have *some* dialups using addresses on the same network as the net0 interface. Can I proxy-arp some dialups and route the others? I'm wanting to subnet a /24 into 4 * /26's and put the dialups in those. Does the NETserver have any issues using the bottom and top subnets? If so, I'll end up wasting half a /24 :-( Normally I'd just try it and see what gives, but I don't have a spare rack to play with :-( Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: (usr-tc) Rumors for new TC Software?
From: lee@gwinnett.com
Date: 1998-02-01 16:41:45
Anyone heard any rumors to when new Total Control Software may be available? I did download and install the NMC software that was posted last week on the Totalservice web site. However, I am hoping the disconnect problems many of us are plagued with will be addressed, as well as the Quake ping problem. Thanks in advance. Lee
Subject: Re: (usr-tc) HyperARC/Authentication
From: Brian <signal@shreve.net>
Date: 1998-02-01 16:52:31
On Sun, 1 Feb 1998, jcondon wrote: > Brian wrote: > > > Are most of you running PAP only, and is it because you found that > > CHAP/PAP (EITHER) doesnt work nicely? > > Same thing on the Netserver I modem Plus and apperently USR knows about > it because when we called with some radius questions the first thing > they did was have us set it for PAP instead of either. Why this needed > I am not sure but, we tried either after we got it working and it > doesn't work that way. Why either is an option when it doesn't apper to > be possible to make it work makes little sense to me. Maybe it would > work with local users but, not radius? Or only with USR radius? Might shave a NAK or two off negotiation if it did work, since I beleive most stacks negotiate CHAP before PAP. In any case, they shouldnt have "EITHER" as the default if it doesnt work properly. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: RE: (usr-tc) Analog MLPPP
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-01 17:06:49
On Sun, 1 Feb 1998, Marshall Morgan wrote: > On Sunday, February 01, 1998 12:35 PM, Brian [SMTP:signal@shreve.net] wrote: > > On Sun, 1 Feb 1998, System Administrator wrote: > > > > > On Sun, 1 Feb 1998, James Wilson wrote: > > > > > > > Anybody have any luck with analog mulitlink PPP. Is this even supported > > > > in > > > > the TC hub? I've seen reports of it being used in Ascend boxes. I tried > > > > it > > > > with two Sportsters and right after authentication, when the MLPPP > > > > negotiation starts, my computer reboots. This is with DUN 1.2. I have a > > > > client that is really interested in this. > > > > > > Yes, lots o' luck here. We have multiple customers using DUN 1.2 w/ two > > > modems to establish MLPPP. Our official support policy on analog MLPPP is > > > "hands-off". In other words, "Yes, we know it will work, but it's up to > > > you to figure out how to make it work." > > > > But you do go in there and increase the port limit for that user to 2 > > channels right? Or does your default RADIUS entry allow for 2 ports to > > begin with? > > I assume customers are charged more for MLPPP right? Yes. Our radius server actually uses "Port-Limit" internally as well as sending it as a response back to the NAS. The radius server constantly monitors all connections, and will disallow authentication attempts if the given user is already connected the "max" number of times (typically one, but not always). Thus, a customer can purchase "x" number of simultaneous logins. If they want to use these from multiple dial-up systems or if they want to do MLPPP, we look at it as the same thing on our side. Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Analog MLPPP
From: James Wilson <james@cruxnet.com>
Date: 1998-02-01 17:15:39
> From: Marshall Morgan <marshall@netdoor.com> > >On Sunday, February 01, 1998 12:35 PM, Brian [SMTP:signal@shreve.net] wrote: >> On Sun, 1 Feb 1998, System Administrator wrote: >> >> > On Sun, 1 Feb 1998, James Wilson wrote: >> > >> > > Anybody have any luck with analog mulitlink PPP. Is this even supported >> > >> > Yes, lots o' luck here. We have multiple customers using DUN 1.2 w/ two >> >> But you do go in there and increase the port limit for that user to 2 >> channels right? Or does your default RADIUS entry allow for 2 ports to >> begin with? > >I assume customers are charged more for MLPPP right? > >Marshall Morgan Yep, we charge them 2 dialup accounts and a static IP address. James
Subject: (usr-tc) Archives updated
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-01 17:25:25
I updated the archives of this list that are stored at http://usr-tc.datasys.net/ This is a webified version of the archives at ftp://ftp.xmission.com/pub/lists/usr-tc/archive/ They're indexed by WMF by thread, subject, sender, and date, and also regular-expression searchable, by WN. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: (usr-tc) List lag?
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-01 19:19:27
Is it just me, or is everyone seeing huge lag on this mailing list? It sometimes takes up to 2 days to get my own posts back. Other (much larger) lists I'm on typically turn the messages around in a matter of minutes. Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: (usr-tc) Using quad modems with serial ports
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-01 19:27:14
My PRI NAC has T1 code loaded on it, and is set to autobusy channels. I want to hook some of my quad modems to a Portmaster through serial ports on a NIC card. The problem is, after I deactivate the modem in the Netserver and reset the port, the channels on that card go into busyout status. I got one out of busyout, and working with the serial port, but no idea how. Any suggestions to get this working? Brian
Subject: Re: (usr-tc) Where to Buy, and a Question!
From: Phil Freed <phil@freed.com>
Date: 1998-02-01 20:24:18
> We are disabling our POP and have the associated equipment for sale, to > include a TC hub w/digital X2 quad modems, etc. I'm a little late - but if you still have this (or other) equipment available, we'd be interested. You can reach one of Gary Seekins Phil Freed Steve Gilbert at our office at 1-800-628-1013. Thanks. --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."
Subject: (usr-tc) TC for POP backup??
From: Henry Moats <nc0419@corp.netcom.com>
Date: 1998-02-02 11:54:55
Has anyone every contemplated using a TC unit to backup a POP site? ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________|
Subject: (usr-tc) Rocket modems & USR TC
From: Jim Faulkner II <jlf@montrose-colo.com>
Date: 1998-02-02 12:00:08
Hello All, I am considering going with 8 port rocket modems in addition to my TC hub. Has anyone done this? Can they be configured to work with USR's TC Radius? Does anybody have any other ideas for expansion short of a new rack and or the new 24 port ($8,000) USR modem cards? Thank You Jim Faulkner
Subject: (usr-tc) Delay in posting to the list
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 12:09:02
Both Bob and Mike are subscribed to different addresses than what they send from. Since Majordomo doesn't find their sender's address on the list, it will bounce to me for approval, thus the delay. The fix is to either subscribe the sending address, or let me know this is impossible. I have a work-around, but I'm afraid I'm far to lazy to do it without prompting. ;-) I just wish that everyone's mailers would send out the same address that they receive to.
Subject: Re: (usr-tc) List lag?
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-02 12:41:54
> Is it just me, or is everyone seeing huge lag on this mailing list? It > sometimes takes up to 2 days to get my own posts back. > > Other (much larger) lists I'm on typically turn the messages around in a > matter of minutes. Haven't seen it here... Usually only takes a few minutes to see my responces bounce back to me. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) (USR-TC) HIPERARC PROBLEM
From: G. Turner <turners@best.com>
Date: 1998-02-02 12:42:55
Just how does the NetServer do MPIP? Could you explain briefly how it does it? Someone told me that it requires a seperate registration server. Let's say I have two chassis in one POP, could one of the NetServer's be the 'registration server' or do I need a unix or Windows server? Next, how will HiPer ARC do MPIP? Will it work with NetServer or will ARC do MPIP in a completely different way? Oh, and will there be a training class on HiPer offered soon? When I asked, someone told me there wasn't enough demand for it yet so limited classes are available. At 03:15 AM 1/29/98 -0600, Tatai SV Krishnan wrote: >Yes, soon > >krish > >On Thu, 29 Jan 1998, Pete Ashdown wrote: > >> Tatai SV Krishnan said once upon a time: >> >> >HiPER ARC and MPIP? HiPer ARC currently does not support MPIP. What is >> >the problem exactly? ARe you talking about MPIP or HiPerDSP? >> >> Are we ever going to see MPIP or an MPIP-like solution for HiPer ARC? >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the message. >> For information on digests or retrieving files and old messages send >> "help" to the same address. Do not use quotes in your message. >> > >----------------------------------------- > \ T.S.V. Krishnan \ > \ Network System Engineer \ ( : - : ) > \ 3Com ............ \ > ----------------------------------------------/ >tkrishna@bubba.ae.usr.com >----------------------------/ http://interproc.ae.usr.com ----/ >-------------------------------------------------------------------------\ > Any Sufficiently advanced bug is indistinguishable for a feature. > - Rick Kulawiec >-------------------------------------------------------------------------/ > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > ++++++++++++++++++++++++ mailto:turners@best.com ++++++++++++++++++++++++
Subject: (usr-tc) X2 and Cellular options...
From: G. Turner <turners@best.com>
Date: 1998-02-02 12:52:53
Hi, Someone I was talking to thought that you have to pay for X2 support on the HiPer DSP. I looked on the 3Com site and it looks to me like X2 is included. Right? What about cellular support on DSP? I suppose they were confusing the DSP with the modem cards. You do have to pay for X2 and Cellular on those, right? ++++++++++++++++++++++++ mailto:turners@best.com ++++++++++++++++++++++++
Subject: Re: (usr-tc) List lag?
From: Mike <mike@coredump.ae.usr.com>
Date: 1998-02-02 13:04:59
At 07:19 PM 2/1/98 +1100, you wrote: > >Is it just me, or is everyone seeing huge lag on this mailing list? It >sometimes takes up to 2 days to get my own posts back. > >Other (much larger) lists I'm on typically turn the messages around in a >matter of minutes. > I'm noticing about 24 hr delay from post to seeing my message on the list.. Pete, are you using the latest MAJORDOMO? -m
Subject: Re: (usr-tc) List lag?
From: James Wilson <james@cruxnet.com>
Date: 1998-02-02 13:59:52
> From: Bob Purdon <bobp@southcom.com.au> > >Is it just me, or is everyone seeing huge lag on this mailing list? It >sometimes takes up to 2 days to get my own posts back. > >Other (much larger) lists I'm on typically turn the messages around in a >matter of minutes. > >Regards, > >Bob Purdon, >Technical Manager, Nope, I get mine back in a couple of minutes. James
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Charles Sprickman <spork@inch.com>
Date: 1998-02-02 14:08:14
I have something of a follow-on question... We're currently using an older Merit Radius server (2.23.24?) which works, but I feel like I'm missing out on some goodies. I'm looking to either upgrade to the Livingston code or the newest Merit release, but I'm a bit confused about the vendor-specific attributes and how one goes about "merging" the USR dictionary in. Also, does any (free) Radius server support port-limits? We have source to the USR radiusd (don't ask) but it supposedly requires some external pay-to-play stuff to actually compile... Thanks, Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Sat, 31 Jan 1998, Mike Hamrich wrote: > Date: Sat, 31 Jan 1998 10:53:16 -0500 > From: Mike Hamrich <mhamrich@drfast.net> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@mail.xmission.com > Subject: (usr-tc) Help ! Radius not accurate > > We are testing Radius 2.01 accouting server. The logs are missing many > events. We have set acc on and are pointing to our Linux Radius. > > Is there some "hidden" setting to make the netserver put out normal no USR > specific logs > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 14:49:24
Charles Sprickman said once upon a time: >I have something of a follow-on question... We're currently using an >older Merit Radius server (2.23.24?) which works, but I feel like I'm >missing out on some goodies. I'm looking to either upgrade to the >Livingston code or the newest Merit release, but I'm a bit confused about >the vendor-specific attributes and how one goes about "merging" the USR >dictionary in. Someone posted a patch a month or so ago. I haven't implemented it yet, so I can't comment on it. >Also, does any (free) Radius server support port-limits? We have source >to the USR radiusd (don't ask) but it supposedly requires some external >pay-to-play stuff to actually compile... The Merit does, use "Port-Limit". This only applies to single chassis though.
Subject: (usr-tc) TCS 3.0.2 - NEW MODEM CODE!!
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 14:51:30
I went looking for documentation for an MP16 and was delighted to find that the new modem code for quads was released today! Start the stampede over to Totalservice.
Subject: Re: (usr-tc) List lag?
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:00:50
On Sun, 1 Feb 1998, Bob Purdon wrote: > > Is it just me, or is everyone seeing huge lag on this mailing list? It I don't know. You running this mail list thru a Netserver card pete? <eg> <duck!> > > Regards, > > Bob Purdon, > Technical Manager, > Southern Internet Services. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NETServer, routing, proxy-arp, subnet-0
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:06:00
On Sun, 1 Feb 1998, Bob Purdon wrote: > > Hi All, > > After a bit of advice here... > > Currently I have NETServer cards at 3.6.28 and 3.5.33, the ptpX and net0 > interfaces sit on the same network, and we use proxy-arp. > > I'm wanting to move the ptpXX interfaces to another network and route the > traffic there - I know that bit works. > > In addition, I need to have *some* dialups using addresses on the same > network as the net0 interface. Can I proxy-arp some dialups and route the > others? I have been there and it aint pretty :). If you *can*, try and get some more address space, and put your net0 interfaces on the same network as your normal LAN at your business. put the ptpx interfaces on there own network(s). We have a seperate network for static IP numbers than our dynamic pools. We route /26's to all of our hubs. I ran into some serious problems of our Cisco adding an ARP entry for a user while they were dialed into TC-HUB1, and then the user hang's up, and dials into TC-HUB2 and can't get anywhere, because the Cisco's arp cache didn't expire. You can do funky things like expire the arp cache every minute or so (yuk!), other than that you can run into some problems. More address space, thats what I would do. > > I'm wanting to subnet a /24 into 4 * /26's and put the dialups in those. > Does the NETserver have any issues using the bottom and top subnets? If > so, I'll end up wasting half a /24 :-( No, the Netserver is CIDR aware. Unlike back in the days of "classfull" you *CAN* keep your upper and lower subnets today. > > Normally I'd just try it and see what gives, but I don't have a spare rack > to play with :-( > > Regards, > > Bob Purdon, > Technical Manager, > Southern Internet Services. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) TC for POP backup??
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 15:06:19
Henry Moats said once upon a time: > > >Has anyone every contemplated using a TC unit to backup a POP site? To backup a POP site (scratches head)? I use them at POP sites. Can you be more specific?
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:07:55
On Mon, 2 Feb 1998, Charles Sprickman wrote: > I have something of a follow-on question... We're currently using an > older Merit Radius server (2.23.24?) which works, but I feel like I'm > missing out on some goodies. I'm looking to either upgrade to the > Livingston code or the newest Merit release, but I'm a bit confused about > the vendor-specific attributes and how one goes about "merging" the USR > dictionary in. > > Also, does any (free) Radius server support port-limits? We have source > to the USR radiusd (don't ask) but it supposedly requires some external > pay-to-play stuff to actually compile... Port-Limit is supported by Merit. But the functionality your looking for you would have to go with a third party product like "tsmon" (aka PMMON). I emailed Brad Owens, the author, today, and he is getting TSMON to work with the new HyperARC equiptment as well. Brian > > Thanks, > > Charles > > ~~~~~~~~~ ~~~~~~~~~~~ > Charles Sprickman Internet Channel > INCH System Administration Team (212)243-5200 > spork@inch.com access@inch.com > > On Sat, 31 Jan 1998, Mike Hamrich wrote: > > > Date: Sat, 31 Jan 1998 10:53:16 -0500 > > From: Mike Hamrich <mhamrich@drfast.net> > > Reply-To: usr-tc@lists.xmission.com > > To: usr-tc@mail.xmission.com > > Subject: (usr-tc) Help ! Radius not accurate > > > > We are testing Radius 2.01 accouting server. The logs are missing many > > events. We have set acc on and are pointing to our Linux Radius. > > > > Is there some "hidden" setting to make the netserver put out normal no USR > > specific logs > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) (USR-TC) HIPERARC PROBLEM
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:10:42
On Mon, 2 Feb 1998, G. Turner wrote: > > Just how does the NetServer do MPIP? Could you explain briefly how it does > it? Someone told me that it requires a seperate registration server. > Let's say I have two chassis in one POP, could one of the NetServer's be > the 'registration server' or do I need a unix or Windows server? Yes. Each chassis that wants to be able to do MPIP must be a MPIP client. This is no problem. Then you must have at least 1 MPIP "server". The server code is built into ComOS, you merely have to enable it. You tell each of your USR hubs the IP address of the server. Then on the server you tell it the IP addresses of the clients. This way, there is always a server that knows about all MPP connections on all hubs. > > Next, how will HiPer ARC do MPIP? Will it work with NetServer or will ARC > do MPIP in a completely different way? Would be nice if it does work with the Netserver's MPIP implementation. > > Oh, and will there be a training class on HiPer offered soon? When I > asked, someone told me there wasn't enough demand for it yet so limited > classes are available. > > > At 03:15 AM 1/29/98 -0600, Tatai SV Krishnan wrote: > >Yes, soon > > > >krish > > > >On Thu, 29 Jan 1998, Pete Ashdown wrote: > > > >> Tatai SV Krishnan said once upon a time: > >> > >> >HiPER ARC and MPIP? HiPer ARC currently does not support MPIP. What is > >> >the problem exactly? ARe you talking about MPIP or HiPerDSP? > >> > >> Are we ever going to see MPIP or an MPIP-like solution for HiPer ARC? > >> > >> - > >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > >> with "unsubscribe usr-tc" in the body of the message. > >> For information on digests or retrieving files and old messages send > >> "help" to the same address. Do not use quotes in your message. > >> > > > >----------------------------------------- > > \ T.S.V. Krishnan \ > > \ Network System Engineer \ ( : - : ) > > \ 3Com ............ \ > > ----------------------------------------------/ > >tkrishna@bubba.ae.usr.com > >----------------------------/ http://interproc.ae.usr.com ----/ > >-------------------------------------------------------------------------\ > > Any Sufficiently advanced bug is indistinguishable for a feature. > > - Rick Kulawiec > >-------------------------------------------------------------------------/ > > > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > > ++++++++++++++++++++++++ > mailto:turners@best.com > ++++++++++++++++++++++++ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) New HDM code
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 15:13:59
It looks as if there is new HDM code as well, version 1.1.1. The date seems to be wrong on the web page. I'm trying to find some release notes for it as well.
Subject: (usr-tc) HP OpenView/TotalControl
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:14:54
We were recently looking into complete network SNMP Managment packages, and inquired about HP OpenView since it seems to be such a well known package. Our sales rep told us there is a "HP OpenView/Total Control", which is for the USR Total Control gear. Does anyone know anything about this? is it good? price? Also, in regards to HPOV in general, is it worth it? What are some packages you all are using/recommend. We have a network that would include: linux running snmpd USR Total Control Gear Cisco Catalyst switches Cisco Routers Ascend Pipelines Office Routers (compatible, cayman, livingston) 3com SuperStack Hub APC UPS's with snmp modules We would really like to manage all this gear from a good interface. Capibilities are Linux, Solaris, NT, Mac, Win95, so any of those platforms would do, I lean toward UNIX or NT though. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Java for USRTC
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:17:36
I think that Java is "ideal" for applications such as TCM. There are lots of programs, such as TCM, where scalability demands outweigh perforamnce. The performance of Java isn't that bad, and the "write once" idea I really like. Anyone know if USR sees the light and is thinking of making Java managment applications for its Total Control gear? I also think some products, could benefit by having small Webservers built into them (kinda like cisco), where you just http to there IP, and bring up a managment frontend, in Java of course. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) 3.0.2 no Netserver code?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 15:18:18
It looks as if there is currently no Netserver code on the TotalService site. I presume this is coming shortly. Here's the notice they have up about 3.0.2: TCS 3.0.2 Release Information The following products have been released as part of the TCS 3.0.2 System Maintenance Release: Quad Modem Double-Sided v5.7.6 Quad Modem Single-Sided v5.8.6 EdgeServer v1.6.0 (includes Virtual Motion Remote Access Manager v1.3) Dual E1/CAS v1.2.9 Refer to the individual release notes for details on what is included in each release. TCS 3.0.2 has been released for use with EdgeServer-based systems only. There will be new NETServer code available soon that will be compatible with the other components of TCS 3.0.2.
Subject: Re: (usr-tc) X2 and Cellular options...
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:20:09
On Mon, 2 Feb 1998, G. Turner wrote: > Hi, > > Someone I was talking to thought that you have to pay for X2 support on the > HiPer DSP. I looked on the 3Com site and it looks to me like X2 is > included. Right? What about cellular support on DSP? x2 is included. Cellular is not. From Appendix L of the HyperDSP owners manual: Quad features not supported by Hiper DSP v1.0 cellular protocols fax protocols v.fc, v.23, v.25 bis, v.54 and HST flow control (software OR hardware) loopback testing online command mode remote access remote password synchronous support > > I suppose they were confusing the DSP with the modem cards. You do have to > pay for X2 and Cellular on those, right? > > > > ++++++++++++++++++++++++ > mailto:turners@best.com > ++++++++++++++++++++++++ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) short haul nic on PRI
From: Brian <signal@shreve.net>
Date: 1998-02-02 15:22:24
is the short haul nic used on the Dual PRI card at all? I have tried to configure the settings for it and it just says "N/A". We just had 2 new spans installed. The cable drop for these spans is routed differently than our other spans, it has a 500ft cable drop! So I was hoping to set the short haul distance for the appropriate length. Can you do this or is it automagically being done already? Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) TCS 3.0.2 - NEW MODEM CODE!!
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-02 16:12:10
> I went looking for documentation for an MP16 and was delighted to find that > the new modem code for quads was released today! > > Start the stampede over to Totalservice. I don't know how true the statement is, but there is a note on the page that components of 3.0.2 are for EDGESERVER chassis only, and there will be a release for netserver based systems later. Just a FYI! | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: (usr-tc) RJ-45 serial pins?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-02 16:27:18
I'm looking for the pin assignments for the RJ-45 serial ports on the back of an MP16. This is the same pin assignments as what is used with the console ports on the back of a TC. Does anyone know what these are?
Subject: Re: (usr-tc) New HDM code
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-02 16:29:37
> It looks as if there is new HDM code as well, version 1.1.1. The date > seems to be wrong on the web page. I'm trying to find some release notes > for it as well. Take a closer look.... The 1.1.1 code is for E1 HDM customers, not T1/PRI. Unless of course, you happen to be one of those E1 customers... <g> | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) short haul nic on PRI
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-02 16:43:52
> is the short haul nic used on the Dual PRI card at all? > > I have tried to configure the settings for it and it just says "N/A". > > We just had 2 new spans installed. The cable drop for these spans is > routed differently than our other spans, it has a 500ft cable drop! So I > was hoping to set the short haul distance for the appropriate length. Can > you do this or is it automagically being done already? There is a seperate short-haul NIC you would have to buy, from what I have been told. I wouldn't worry about it - the long haul seems to sync up OK to "hot" signals. Some of my chassis are fed out of my T3 mux that only has about 5ft or less of cable between the port on the mux, and the total control. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) HP OpenView/TotalControl
From: matthew <matthew@the-spa.com>
Date: 1998-02-02 16:46:51
hp will send you a 30 day eval of openview but it is a big program and pretty hard to figure out without docs which the eval does NOT include. also, there is hp openview suite and openview network node manager plus a bunch of specific ones like the usr total control one you mentioned. matthew At 03:14 PM 2/2/98 -0600, Brian wrote: > >We were recently looking into complete network SNMP Managment packages, >and inquired about HP OpenView since it seems to be such a well known >package. Our sales rep told us there is a "HP OpenView/Total Control", >which is for the USR Total Control gear. > >Does anyone know anything about this? is it good? price? > >Also, in regards to HPOV in general, is it worth it? What are some >packages you all are using/recommend. > >We have a network that would include: > >linux running snmpd >USR Total Control Gear >Cisco Catalyst switches >Cisco Routers >Ascend Pipelines >Office Routers (compatible, cayman, livingston) >3com SuperStack Hub >APC UPS's with snmp modules > >We would really like to manage all this gear from a good interface. >Capibilities are Linux, Solaris, NT, Mac, Win95, so any of those platforms >would do, I lean toward UNIX or NT though. > >Brian > > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Systems Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) HP OpenView/TotalControl
From: matthew <matthew@the-spa.com>
Date: 1998-02-02 16:46:51
hp will send you a 30 day eval of openview but it is a big program and pretty hard to figure out without docs which the eval does NOT include. also, there is hp openview suite and openview network node manager plus a bunch of specific ones like the usr total control one you mentioned. matthew At 03:14 PM 2/2/98 -0600, Brian wrote: > >We were recently looking into complete network SNMP Managment packages, >and inquired about HP OpenView since it seems to be such a well known >package. Our sales rep told us there is a "HP OpenView/Total Control", >which is for the USR Total Control gear. > >Does anyone know anything about this? is it good? price? > >Also, in regards to HPOV in general, is it worth it? What are some >packages you all are using/recommend. > >We have a network that would include: > >linux running snmpd >USR Total Control Gear >Cisco Catalyst switches >Cisco Routers >Ascend Pipelines >Office Routers (compatible, cayman, livingston) >3com SuperStack Hub >APC UPS's with snmp modules > >We would really like to manage all this gear from a good interface. >Capibilities are Linux, Solaris, NT, Mac, Win95, so any of those platforms >would do, I lean toward UNIX or NT though. > >Brian > > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Systems Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) short haul nic on PRI
From: buster_joseph/mw/us/3com@usr.com
Date: 1998-02-02 16:49:05
--IMA.Boundary.003064688 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part For the Dual PRI card you will need to have a Short haul NIC to use the short haul setting. The standard NIC only does long haul. Buster Brian <signal@shreve.net> AT Internet on 02/02/98 03:22:24 PM Internet@ccmail cc: --IMA.Boundary.003064688 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part is the short haul nic used on the Dual PRI card at all? I have tried to configure the settings for it and it just says "N/A". We just had 2 new spans installed. The cable drop for these spans is routed differently than our other spans, it has a 500ft cable drop! So I was hoping to set the short haul distance for the appropriate length. Can you do this or is it automagically being done already? Brian /-------------------------- signal@shreve.net | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318) 222-2638 | | Systems Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | \-------------------------- 318-222-2638 x109 - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. --IMA.Boundary.003064688 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4D640230; Mon, 2 Feb 98 15:52:35 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id PAA27749; Mon, 2 Feb 1998 15:09:12 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0xzTK1-0005Wl-00; Mon, 2 Feb 1998 14:22:29 -0700 Received: from mail.xmission.com [198.60.22.22] by lists.xmission.com with smtp (Exim 1.73 #4) id 0xzTJx-0005W2-00; Mon, 2 Feb 1998 14:22:25 -0700 Received: from mercury.shreve.net [208.206.76.23] by mail.xmission.com with esmtp (Exim 1.73 #4) id 0xzTJv-00027e-00; Mon, 2 Feb 1998 14:22:23 -0700 Received: from localhost (localhost [127.0.0.1]) by mercury.shreve.net (8.8.7/8.8.7) with SMTP id PAA07308 for <usr-tc@xmission.com>; Mon, 2 Feb 1998 15:22:24 -0600 Message-ID: <Pine.LNX.3.96.980202152024.5365L-100000@mercury.shreve.net> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.003064688--
Subject: Re: (usr-tc) RJ-45 serial pins?
From: Charles Hill <chill@ionet.net>
Date: 1998-02-02 17:55:34
They are EIA/TIA-561 standard RJ-45 RS-232 serial jacks, so the pins are like this: EIA/TIA-561 "Simple 8 Position Non-Syncronous Interface Between Data Terminal Equipment and Data Circuit-Terminating Equipment Employing Data Interchange" Dated: October 1990 pin signal
Subject: Re: (usr-tc) Java for USRTC
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-02 18:04:05
On Mon, 2 Feb 1998, Brian wrote: > I think that Java is "ideal" for applications such as TCM. There are lots > of programs, such as TCM, where scalability demands outweigh perforamnce. > The performance of Java isn't that bad, and the "write once" idea I really > like. > > Anyone know if USR sees the light and is thinking of making Java managment > applications for its Total Control gear? > > I also think some products, could benefit by having small Webservers built > into them (kinda like cisco), where you just http to there IP, and bring > up a managment frontend, in Java of course. There are web servers built into a few 3Com products. The ADSL products like the Viper or Axcell card both have it and the Lan-linkers as well. It is basically just netstats and routes and information about the unit. It is basic, but it is a start. (: :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) HyperARC/Authentication
From: MegaZone <megazone@livingston.com>
Date: 1998-02-02 20:37:39
Once upon a time Mike Wronski shaped the electrons to say... >CHAP being more secure is usually negotiated first.. And in the LCP RFC >MSCHAP & CHAP cant be differentiated (90% positive on this fact, I will >have to investigate).. Once an auth proto is chosen and then fails, Yes they can be. Off the top of my head: RFC-CHAP shows as 03 05 C2 23 05 MS--CHAP shows as 03 05 C2 23 80 (PAP is 03 04 C0 23) We have a similar problem. If a windows box insists on MS-CHAP we would not recognize what they were asking for and we'd fight back and forth until we gave up and failed. The plan is to recognize MS-CHAP and deliberately counter request RFC CHAP and/or PAP. It only happens once in a while, I eventually debugged the race condition and opened the engineering issue on this which is one more reason I loath MS 'extensions'. -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: Re: (usr-tc) HyperARC/Authentication
From: Mike Wronski <mwronski@coredump.ae.usr.com>
Date: 1998-02-02 22:26:20
On Mon, 2 Feb 1998, Jeff Mcadams wrote: > Thus spake jcondon > >Mike Wronski wrote: > >> The issue hear is with Microsoft. Their implementation of CHAP (MSCHAP) > >> is getting negotiated. But they use a different encryption hash (MD4 I > >> think) causing the authentication to fail. > > >Figures. So if MSChap fails in WIN95 it doesn't revert back to PAP and > >complete the call? > > Well, I believe the choice of which authentication to use is negotiated > during LCP, so they would have to go back and renegotiate LCP to choose > another auth protocol. CHAP being more secure is usually negotiated first.. And in the LCP RFC MSCHAP & CHAP cant be differentiated (90% positive on this fact, I will have to investigate).. Once an auth proto is chosen and then fails, there is usually not another CONF_REQ for a different proto. The call will usually drop which is the case of Win95 -> ARC/NS. This is, however now fixed in the ARC. I think the fix is in the GA (4.0.19) code.. My thoughts earlier were on a older BETA code that was broke. -m
Subject: Re: (usr-tc) HyperARC/Authentication
From: jcondon <jcondon@computer.net>
Date: 1998-02-02 22:48:12
Mike Wronski wrote: > > The issue hear is with Microsoft. Their implementation of CHAP (MSCHAP) > is getting negotiated. But they use a different encryption hash (MD4 I > think) causing the authentication to fail. > Figures. So if MSChap fails in WIN95 it doesn't revert back to PAP and complete the call? If I call in with Unix box with both enabled Chap would work then since it isn't the braindead MS version? Now that you mention it I seem to remeber reading that MSChap wasn't compatible with the real world.
Subject: Re: (usr-tc) HyperARC/Authentication
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-02 23:07:52
Thus spake jcondon >Mike Wronski wrote: >> The issue hear is with Microsoft. Their implementation of CHAP (MSCHAP) >> is getting negotiated. But they use a different encryption hash (MD4 I >> think) causing the authentication to fail. >Figures. So if MSChap fails in WIN95 it doesn't revert back to PAP and >complete the call? Well, I believe the choice of which authentication to use is negotiated during LCP, so they would have to go back and renegotiate LCP to choose another auth protocol. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) HyperARC/Authentication
From: jcondon <jcondon@computer.net>
Date: 1998-02-03 00:41:59
Mike Wronski wrote: > > On Mon, 2 Feb 1998, Jeff Mcadams wrote: > > This is, however now fixed in the ARC. I think the fix is in the GA > (4.0.19) code.. My thoughts earlier were on a older BETA code that was > broke. We however have a Netserver and not an ARC.
Subject: Re: (usr-tc) HyperARC/Authentication
From: jcondon <jcondon@computer.net>
Date: 1998-02-03 00:47:37
MegaZone wrote: > > It only happens once in a while, I eventually debugged the race condition > and opened the engineering issue on this which is one more reason I loath > MS 'extensions'. > I loath MS completely. For this and about 3 or 4 trillion other reasons. Why they can't go with the flow on established standards is beyond me. Instead the dream up some half baked ways of doing things that is incompatible with the rest of the world. Are they hoping that their less secure MD4 way will win over the rest of the internet? Do they get a license fee if others use their MSChap scheme? Half baked Java. 1/3 baked implementation of TCP/IP and DNS. I could go on forever. But, you knew this already.
Subject: (usr-tc) Merrit Radius
From: Terry Kennedy <terry@olypen.com>
Date: 1998-02-03 09:47:01
Thanks for reply Steve, We are using SCO Open Server and a Win95 box. We use merrit Radius on the SCO Servers for authentication and I use the Win95 box to configure the hardware.The people at Merrit tell us that we have to merge the USR dictionary with thier's and that this is up to us. We can figure that out, but I have a couple of question if you don't mind. 1) Where do I get the Dictionary? At one time I had downloaded the Solaris version of your Radius server but have dicarded this. Those files are now locked to me from totalservice.usr.com, I told them I would be using the Win95 /NT version. The only file that I can find in the Win95/NT version is a file called accdict.dat. Is this the file we are looking to merge our Merrit dictionary with? This quote is from the file. It would appear to me that we need the Security end of this, am I right or way off track? This is a pared down version of the Security and Accounting V5.0 # Dictionary, created on Oct 28,1997. It is to be used by the # Accounting server, not by the Security And Accounting Server. 2) What is the rest of the sane world using? Are they all really running Sun Servers and Solaris? I am not looking for USR to configure our systems for us, just a point in right direction. Any help in this matter would greatly appreciated. Thanks much Terry Kennedy - > > You can find this in your dictnary.dat or documentation. The > format is four bytes of parameter followed by four bytes of > value, for example, Vendor-Specific=vUSR-00009023 0000000c = > parm x'9023' (connect speed) , value x'c' (21,600 bps). > You can control accounting info from the NMC's 'log group > selection'. > > > Using Merrit Radius these show up in our logs. Does anyone know > how we can identify these? I have looked in the reference I > recieved with my TC's but can't find a match. What is the format > of the numbers at the end? Can you control what values the TC's > send to Radius? > > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-000090230000000c > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-0000006c00000013 > non-enapsulated vendor specific attribute > Vendor-Specific=vUSR-0000009900000003 > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-000000c700000001 > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-0000901900000000 > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-0000901a00000000 > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-0000901b00000000 > non-encapsulated vendor specific attribute > Vendor-Specific=vUSR-0000901d00000013 > >
Subject: Re: (usr-tc) X2 and Cellular options...
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 10:00:47
On Mon, 2 Feb 1998, Brian wrote: > >From Appendix L of the HyperDSP owners manual: > > Quad features not supported by Hiper DSP v1.0 > --------------------------------------------- > > cellular protocols > fax protocols > v.fc, v.23, v.25 bis, v.54 and HST > flow control (software OR hardware) How the heck do these modems even work without flow control? Won't they just keep sending data forever, even if the other end is telling them to pause for a moment while the other end catches up? Brian
Subject: (usr-tc) Maximum ports for MPPP modems/ISDN?
From: G. Turner <turners@best.com>
Date: 1998-02-03 10:31:09
What is the maximum number of ports attainable today using Total Control in an MPPP solution per chassis? Assuming PRI only... Here's the combinations I came up with: Ports Routers Modems/ISDN Type 230 5 NetServer 10 HiPer DSP Modem/ISDN 161 1 ARC 7 HiPer DSP Modem/ISDN 94 1 NetServer 2 HiPer DSP Modem/ISDN 1 Dual PRI 12 Quad 47 1 NetServer Modem/ISDN 1 Dual PRI 12 Quad Am I right? I wish 3Com's web site had some kind of matrix showing all the possible combinations. It's a pain trying to figure it out unless you're a wizard at this already. What about other combinations using T1? What is the maximum you folks are using right now and what components did you use to get to that capacity? I need to be able to use modems and ISDN. This kinda info would be great for a FAQ! ++++++++++++++++++++++++ mailto:turners@best.com ++++++++++++++++++++++++
Subject: (usr-tc) HiPer ARC, redundancy?
From: G. Turner <turners@best.com>
Date: 1998-02-03 10:35:08
If I have two ARCs in one chassis will one take over if the other fails? How do I set that up? I've heard it said that this can be done but I couldn't find it on the docs. ++++++++++++++++++++++++ mailto:turners@best.com ++++++++++++++++++++++++
Subject: (usr-tc) TCS Dial out
From: Randy Doran <rtdoran@gate.net>
Date: 1998-02-03 10:35:36
Is anybody using the I-Modems on a TCS Hub to dial out? I would like to set up a chassis to dial our dedicated ISDN customers so they don't have to dial us, thus saving on the Bell South BRI tarriff. We can make our Ascend Max dial the customer and when the connection is lost it attempts to reconnect automatically. I suppose dial back would be OK but I would prefer it to be automatic to nail up the circuit. Thanks, | Randy Doran CyberGate Network Operations | | Circuit Engineer\ 1301 W Newport Center Dr. | | Modem Network Administrator Deerfield Beach, FL 33442 | | 954-429-8069 FAX 954-429-8001 |
Subject: Re: (usr-tc) X2 and Cellular options...
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-03 10:52:49
Brian Elfert said once upon a time: >> Quad features not supported by Hiper DSP v1.0 >> --------------------------------------------- >> flow control (software OR hardware) > >How the heck do these modems even work without flow control? > >Won't they just keep sending data forever, even if the other end is >telling them to pause for a moment while the other end catches up? The packet bus handles it. The fact that Quads have it relates to the fact that some Quads have serial ports.
Subject: Re: (usr-tc) TCS Dial out
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-03 10:55:20
Randy Doran said once upon a time: > >Is anybody using the I-Modems on a TCS Hub to dial out? I would like to >set up a chassis to dial our dedicated ISDN customers so they don't have >to dial us, thus saving on the Bell South BRI tarriff. We can make our >Ascend Max dial the customer and when the connection is lost it attempts >to reconnect automatically. I suppose dial back would be OK but I would >prefer it to be automatic to nail up the circuit. Callback on ANI number was something that was promised to me by USR in the winter of '96 when I was evaluating chassis. I have no idea what the current status is.
Subject: (usr-tc) Problems with Couriers
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-03 11:09:38
We recently had a rash of problems with Sportster modems connecting to our TC Hubs. In two of the cases, the modems would handshake and never give a connected message, in several other cases the modem would get garbage when trying to connect with anything higher than a 19.2 port speed. This sounds like an X2 negotiation problem, but it should negotiate down to a v.34 connection instead of giving garbage right? I had one of these people connect to another provider who does not have X2 and they connected at 26.4. Could this be a problem with our boxes? If so, any ideas on how to get it negotiating a better connection? If not, can you turn off X2 negotiation on a Sportster? Thanks for any insights. Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com
Subject: Re: (usr-tc) Problems with Couriers
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-03 11:11:03
Thus spake Peter D. Mayer >We recently had a rash of problems with Sportster modems connecting to our >TC Hubs. In two of the cases, the modems would handshake and never give a >connected message, in several other cases the modem would get garbage when >trying to connect with anything higher than a 19.2 port speed. >This sounds like an X2 negotiation problem, but it should negotiate down to >a v.34 connection instead of giving garbage right? I had one of these >people connect to another provider who does not have X2 and they connected >at 26.4. Could this be a problem with our boxes? If so, any ideas on how >to get it negotiating a better connection? If not, can you turn off X2 >negotiation on a Sportster? Hrmm...sounds vaguely familiar with the symptoms I had a customer describe to me the other day...I had assumed it was a telco problem (based on some other information...telco problems have been legion around here recently...BellSouth doesn't want to play nice with ICG very well), but since you're experiencing something similar, perhaps I was wrong. Did you get what version of code your customers are using on their modem? I've heard and had some customers experience some problems with the sportster code dated 10/3/1997 and 10/6/1997. I neglected to check that with my customer, but will be going back to do that. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) vendor specific attributes
From: steve_valiunas/mw/us/3com@usr.com
Date: 1998-02-03 11:13:40
--IMA.Boundary.882625688 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part You can find this in your dictnary.dat or documentation. The format is four bytes of parameter followed by four bytes of value, for example, Vendor-Specific=vUSR-00009023 0000000c = parm x'9023' (connect speed) , value x'c' (21,600 bps). You can control accounting info from the NMC's 'log group selection'. --IMA.Boundary.882625688 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Using Merrit Radius these show up in our logs. Does anyone know how we can identify these? I have looked in the reference I recieved with my TC's but can't find a match. What is the format of the numbers at the end? Can you control what values the TC's send to Radius? non-encapsulated vendor specific attribute Vendor-Specific=vUSR-000090230000000c non-encapsulated vendor specific attribute Vendor-Specific=vUSR-0000006c00000013 non-enapsulated vendor specific attribute Vendor-Specific=vUSR-0000009900000003 non-encapsulated vendor specific attribute Vendor-Specific=vUSR-000000c700000001 non-encapsulated vendor specific attribute Vendor-Specific=vUSR-0000901900000000 non-encapsulated vendor specific attribute Vendor-Specific=vUSR-0000901a00000000 non-encapsulated vendor specific attribute Vendor-Specific=vUSR-0000901b00000000 non-encapsulated vendor specific attribute Vendor-Specific=vUSR-0000901d00000013 Terry Kennedy Olypen, Inc. terry@olypen.com 360 452 9755 - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. --IMA.Boundary.882625688 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4D0B0BA0; Thu, 29 Jan 98 10:39:22 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id KAA26733; Thu, 29 Jan 1998 10:20:41 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0xxwuE-0002t5-00; Thu, 29 Jan 1998 09:33:34 -0700 Received: from mail.olypen.com [208.200.248.2] by lists.xmission.com with smtp (Exim 1.73 #4) id 0xxwuB-0002sS-00; Thu, 29 Jan 1998 09:33:31 -0700 Received: from modems.olypen.com(src addr [208.200.248.9]) (1462 bytes) by mail.olypen.com via sendmail with P\:esmtp/R:inet_hosts/T:smtp (sender: <terry@olypen.com>) id <m0xxwuG-0002RjC@mail.olypen.com> for <usr-tc@lists.xmission.com>; Thu, 29 Jan 1998 08:33:36 -0800 (PST) (Smail-3.2.0.96 1997-Jun-2 #19 built 1997-Jun-30) Message-Id: <m0xxwuG-0002RjC@mail.olypen.com> X-MSMail-Priority: Normal X-Priority: 3 X-Mailer: Microsoft Internet Mail 4.70.1161 MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.882625688--
Subject: Re: (usr-tc) HiPer ARC, redundancy?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-03 11:38:30
G. Turner said once upon a time: > >If I have two ARCs in one chassis will one take over if the other fails? >How do I set that up? I've heard it said that this can be done but I >couldn't find it on the docs. It can't be done under the current code. The only thing that can be done is assigning different HDMs to different ARCs, so instead of one chassis, you get TWO!
Subject: (usr-tc) Problems with Sportsters
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-03 12:42:35
Oops, got the subject wrong the first time. -----Original Message----- We recently had a rash of problems with Sportster modems connecting to our TC Hubs. In two of the cases, the modems would handshake and never give a connected message, in several other cases the modem would get garbage when trying to connect with anything higher than a 19.2 port speed. This sounds like an X2 negotiation problem, but it should negotiate down to a v.34 connection instead of giving garbage right? I had one of these people connect to another provider who does not have X2 and they connected at 26.4. Could this be a problem with our boxes? If so, any ideas on how to get it negotiating a better connection? If not, can you turn off X2 negotiation on a Sportster? Thanks for any insights. Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com
Subject: Re: (usr-tc) Quake disconnects
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-03 13:51:47
Brian Elfert said once upon a time: > >I have a couple of customers complaining that they get disconnected when >playing Quake. > >I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >complained about any lag problems to date. Are they using X2? If so, reduce the top speed with &N.
Subject: (usr-tc) Quake disconnects
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 14:32:48
I have a couple of customers complaining that they get disconnected when playing Quake. I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's complained about any lag problems to date. Brian
Subject: Re: (usr-tc) Problems with Couriers
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 15:09:03
On Tue, 3 Feb 1998, Peter D. Mayer wrote: > We recently had a rash of problems with Sportster modems connecting to our > TC Hubs. In two of the cases, the modems would handshake and never give a > connected message, in several other cases the modem would get garbage when > trying to connect with anything higher than a 19.2 port speed. > > This sounds like an X2 negotiation problem, but it should negotiate down to > a v.34 connection instead of giving garbage right? I had one of these > people connect to another provider who does not have X2 and they connected Check to be sure that their modem isn't running the 10/?/97 code. There are known problems with this code that do things like that. 3COM/USR can supply the customer with a downgrade wizard that installs the previous rev of code. We've had one customer so far have a problem like this, and they fixed it by downgrading the code. Brian
Subject: Re: (usr-tc) Quake disconnects
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-03 15:15:17
Mike said once upon a time: >You may also want to get the 3.6.69 Quake lag fix. Huh? Wha? Details! ;-)
Subject: Re: (usr-tc) Quake disconnects
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-03 15:19:44
At 01:51 PM 2/3/98 -0700, you wrote: >Brian Elfert said once upon a time: >> >>I have a couple of customers complaining that they get disconnected when >>playing Quake. >> >>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >>complained about any lag problems to date. > >Are they using X2? If so, reduce the top speed with &N. > You may also want to get the 3.6.69 Quake lag fix. -m `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)
Subject: Re: (usr-tc) Quake disconnects
From: Beth Montes <bmontes@iland.net>
Date: 1998-02-03 16:23:42
At 03:19 PM 2/3/98 -0600, Mike wrote: >At 01:51 PM 2/3/98 -0700, you wrote: >>Brian Elfert said once upon a time: >>> >>>I have a couple of customers complaining that they get disconnected when >>>playing Quake. >>> >>>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >>>complained about any lag problems to date. >> >>Are they using X2? If so, reduce the top speed with &N. >> >You may also want to get the 3.6.69 Quake lag fix. Yikes! I missed this! Do you have details??? Beth >-m > >`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' >Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics >Network Systems Engineer >PGP: http://coredump.ae.usr.com/pgp (Prefered) > _____________________________________________ Beth Montes bmontes@iland.net 816-827-5111 x21 ComputerLand of Sedalia http://www.c-land.com I-Land Internet Services http://www.iland.net _____________________________________________
Subject: Re: (usr-tc) X2 and Cellular options...
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 17:14:04
On Tue, 3 Feb 1998, Pete Ashdown wrote: > >> Quad features not supported by Hiper DSP v1.0 > >> --------------------------------------------- > > >> flow control (software OR hardware) > > > >How the heck do these modems even work without flow control? > > > >Won't they just keep sending data forever, even if the other end is > >telling them to pause for a moment while the other end catches up? > > The packet bus handles it. The fact that Quads have it relates to the fact > that some Quads have serial ports. Why mention the lack of things that don't even affect anything? Brian
Subject: Re: (usr-tc) X2 and Cellular options...
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 17:14:04
On Tue, 3 Feb 1998, Pete Ashdown wrote: > >> Quad features not supported by Hiper DSP v1.0 > >> --------------------------------------------- > > >> flow control (software OR hardware) > > > >How the heck do these modems even work without flow control? > > > >Won't they just keep sending data forever, even if the other end is > >telling them to pause for a moment while the other end catches up? > > The packet bus handles it. The fact that Quads have it relates to the fact > that some Quads have serial ports. Why mention the lack of things that don't even affect anything? Brian
Subject: Re: (usr-tc) Quake disconnects
From: Brian <signal@shreve.net>
Date: 1998-02-03 17:20:20
On Tue, 3 Feb 1998, Pete Ashdown wrote: > Mike said once upon a time: > > >You may also want to get the 3.6.69 Quake lag fix. > > Huh? Wha? Details! ;-) its an improvement, not a fix, there is still bigtime latency issues even with 3.6.69. Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. Sometimes, you can get down to about 110-120. What do other 128k users get on your quake server when they *arent* going thru a USR hub? even if they are going thru 6 hops of routers to get to your server? usually they get 50-90ms. Definitly issues. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quake disconnects
From: Brian <signal@shreve.net>
Date: 1998-02-03 17:20:20
On Tue, 3 Feb 1998, Pete Ashdown wrote: > Mike said once upon a time: > > >You may also want to get the 3.6.69 Quake lag fix. > > Huh? Wha? Details! ;-) its an improvement, not a fix, there is still bigtime latency issues even with 3.6.69. Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. Sometimes, you can get down to about 110-120. What do other 128k users get on your quake server when they *arent* going thru a USR hub? even if they are going thru 6 hops of routers to get to your server? usually they get 50-90ms. Definitly issues. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quake disconnects
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-03 17:23:02
At 03:15 PM 2/3/98 -0700, Pete Ashdown wrote: >Mike said once upon a time: > >>You may also want to get the 3.6.69 Quake lag fix. > >Huh? Wha? Details! ;-) > Precisely.. Where do we get 3.6.69?? Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) Quake disconnects
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 17:57:24
On Tue, 3 Feb 1998, Pete Ashdown wrote: > Brian Elfert said once upon a time: > > > >I have a couple of customers complaining that they get disconnected when > >playing Quake. > > > >I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's > >complained about any lag problems to date. > > Are they using X2? If so, reduce the top speed with &N. Nope, both 28.8 and x2. Didn't happen with our MP/16s and PM2Es. We're migrating everything off our MP/16s and this is a problem right now. If it was an x2 problem, I'd think they would disconnect doing just anything, not just during quake. I'm actually thinking of putting 24 of my quad modem on a PM25 to see if that fixes it. Brian
Subject: Re: (usr-tc) Quake disconnects
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-03 18:01:45
On Tue, 3 Feb 1998, Mike wrote: > >>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's > >>complained about any lag problems to date. > > > >Are they using X2? If so, reduce the top speed with &N. > > > You may also want to get the 3.6.69 Quake lag fix. Where can I get this? Do I have to call support? Brian
Subject: Re: (usr-tc) Quake disconnects
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-03 18:12:59
On Tue, 3 Feb 1998, Brian Elfert wrote: > I have a couple of customers complaining that they get disconnected when > playing Quake. > > I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's > complained about any lag problems to date. > I remember in the past I talked to some customers with that problem on older code. I could usually fix them by adding &k0 to their init string. They would not get disconnections, but it would be slower. I have not talked to anyone recently with that problem though. The last time we had it was pre-3.5.x code. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) Maximum ports for MPPP modems/ISDN?
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-03 18:40:35
On Tue, 3 Feb 1998, G. Turner wrote: > What is the maximum number of ports attainable today using Total Control in > an MPPP solution per chassis? > > Assuming PRI only... > > Here's the combinations I came up with: > > Ports Routers Modems/ISDN Type > 230 5 NetServer 10 HiPer DSP Modem/ISDN > 161 1 ARC 7 HiPer DSP Modem/ISDN > 94 1 NetServer 2 HiPer DSP Modem/ISDN > 1 Dual PRI 12 Quad > 47 1 NetServer Modem/ISDN > 1 Dual PRI 12 Quad > > Am I right? I wish 3Com's web site had some kind of matrix showing all the > possible > combinations. It's a pain trying to figure it out unless you're a wizard > at this already. > > What about other combinations using T1? > > What is the maximum you folks are using right now and what components did > you use > to get to that capacity? I need to be able to use modems and ISDN. > > This kinda info would be great for a FAQ! I don't think that is right, at least for the netserver DSP combination. You are only using 2 DSPs per netserver, when the limit is 4 per netserver. We are running 8 Hiper DSP cards in one chassis with 2 EPB netserver cards running off a 70 Amp power supply. According to USR, if you are using the 130 Amp power supply, you can push it up to 12 HiperDSP cards and 3 EPB netserver cards. 12*23=276 ports per chassis. If you are going with Hiper ARCs you can get 14 DSPs and 2 ARCs with a 130 Amp power supply. The most you are supposed to have with a 70 Amp power supply is 10 DSPs and 2 ARCs. 14 DSPs would give you 14*23=322 ports per chassis and 10 DSPs would give you 10*23=230 ports per chassis. Hope this helps. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) List lag?
From: Greg Coffey <greg@coffey.com>
Date: 1998-02-03 18:46:35
There are times, more recently, that some of my posts have taken a day or so to show up. At 01:59 PM 2/2/98 -0400, you wrote: >> From: Bob Purdon <bobp@southcom.com.au> >> >>Is it just me, or is everyone seeing huge lag on this mailing list? It >>sometimes takes up to 2 days to get my own posts back. >> >>Other (much larger) lists I'm on typically turn the messages around in a >>matter of minutes. >> >>Regards, >> >>Bob Purdon, >>Technical Manager, > >Nope, I get mine back in a couple of minutes. > >James > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > Thanks, Greg Coffey, CoffeyNet Voice 307-234-5443 307-234-5446 Fax ~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~ 142 S. Center St. US Robotics x2 56k $20 in Casper Casper, WY 82601 Local Internet for Casper, Rawlins, Douglas, www.coffey.com Wheatland, Pinedale, Lander & Lusk, WY ~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~
Subject: Re: (usr-tc) Problems with Couriers
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-03 20:15:07
Do you know where we can get the downgrade wizard? We now have several customers having this problem and it would be nice to have on hand. Thanks, Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- Cc: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > > >On Tue, 3 Feb 1998, Peter D. Mayer wrote: > >> We recently had a rash of problems with Sportster modems connecting to our >> TC Hubs. In two of the cases, the modems would handshake and never give a >> connected message, in several other cases the modem would get garbage when >> trying to connect with anything higher than a 19.2 port speed. >> >> This sounds like an X2 negotiation problem, but it should negotiate down to >> a v.34 connection instead of giving garbage right? I had one of these >> people connect to another provider who does not have X2 and they connected > >Check to be sure that their modem isn't running the 10/?/97 code. > >There are known problems with this code that do things like that. >3COM/USR can supply the customer with a downgrade wizard that installs the >previous rev of code. > >We've had one customer so far have a problem like this, and they fixed it >by downgrading the code. > >Brian > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) List lag; 1200 bps connections
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-03 21:03:00
Yes, this is way off topic; I guess it's meta-discussion of TC equipment. When sendmail is used as the MTA for lists, the messages are typically sent out in a serial fashion. As such, if your address is close to the top of the file of email addresses, then you'll tend to get yours sooner than somebody close to the bottom. And the list managers get to determine the ordering of the addresses in that file, of course. It's typically just the order in which people subscribed. And to justify this posting, I'll relate an experience: my firm moved five TC units over the weekend as we were moving our entire operation. They'd all been operating fine at the old site taking calls from channelized T1s exclusively. Our only change was from loop start to ground start when we moved. Exactly one of those five racks performed normally when remounted and fired back up. I changed the call parameter settings, of course, and expected that to be my only change. But as it turned out, I've had to reflash several modems and fix a _lot_ of configuration information. In one case, ten contiguous quad modems decided to take calls at 1200 bps -only-.
Subject: Re: (usr-tc) Quake disconnects
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-03 23:18:45
At 03:19 PM 2/3/98 -0600, you wrote: >At 01:51 PM 2/3/98 -0700, you wrote: >>Brian Elfert said once upon a time: >>> >>>I have a couple of customers complaining that they get disconnected when >>>playing Quake. >>> >>>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >>>complained about any lag problems to date. >> >>Are they using X2? If so, reduce the top speed with &N. >> >You may also want to get the 3.6.69 Quake lag fix. > This is a fix available via support for Netserver & Quad Modem combinations.. There are many improvements in the code path for UDP traffic.. These improvements have also been pushed into the 3.7.x code that is in BETA now.. We have mixed reports about improvement.. Some claim their lag went away, some see improvement but not gone, and others claim that the improvement is negligible.. If you want to try it, and are experiencing quake lag you can request it through support or you can wait for the general release of the 3.7.x code. -m
Subject: Re: (usr-tc) Quake disconnects
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-04 02:34:14
On Tue, 3 Feb 1998, Mike wrote: > At 03:19 PM 2/3/98 -0600, you wrote: > >At 01:51 PM 2/3/98 -0700, you wrote: > >>Brian Elfert said once upon a time: > >>> > >>>I have a couple of customers complaining that they get disconnected when > >>>playing Quake. > >>> > >>>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's > >>>complained about any lag problems to date. > >> > >>Are they using X2? If so, reduce the top speed with &N. > >> > >You may also want to get the 3.6.69 Quake lag fix. > > > This is a fix available via support for Netserver & Quad Modem > combinations.. There are many improvements in the code path for > UDP traffic.. These improvements have also been pushed into the > 3.7.x code that is in BETA now.. > > We have mixed reports about improvement.. Some claim their lag went away, > some see improvement but not gone, and others claim that the improvement > is negligible.. If you want to try it, and are experiencing quake lag you > can request it through support or you can wait for the general release of > the 3.7.x code. > Ok.. I am requesting it!!! Please let me have it. Thank you... -Garry
Subject: (usr-tc) Can I upgrade this puppy?
From: G. Turner <turners@best.com>
Date: 1998-02-04 10:00:24
Command> ver U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.2.22 Build date: May 13 1996 Build time: 16:32:10 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Standard Licensed for 60 ports. Can I upgrade this particular NetServer to support the HiPer DSP? Packet Bus says Standard and I thought it might have to be Enhanced to support HiPer DSP? Thanks. ++++++++++++++++++++++++ mailto:turners@best.com ++++++++++++++++++++++++
Subject: Re: (usr-tc) Quake disconnects
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-04 10:22:07
At 02:34 AM 2/4/98 -0500, you wrote: > > >On Tue, 3 Feb 1998, Mike wrote: > >> At 03:19 PM 2/3/98 -0600, you wrote: >> >At 01:51 PM 2/3/98 -0700, you wrote: >> >>Brian Elfert said once upon a time: >> >>> >> >>>I have a couple of customers complaining that they get disconnected when >> >>>playing Quake. >> >>> >> >>>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >> >>>complained about any lag problems to date. >> >> >> >>Are they using X2? If so, reduce the top speed with &N. >> >> >> >You may also want to get the 3.6.69 Quake lag fix. >> > >> This is a fix available via support for Netserver & Quad Modem >> combinations.. There are many improvements in the code path for >> UDP traffic.. These improvements have also been pushed into the >> 3.7.x code that is in BETA now.. >> >> We have mixed reports about improvement.. Some claim their lag went away, >> some see improvement but not gone, and others claim that the improvement >> is negligible.. If you want to try it, and are experiencing quake lag you >> can request it through support or you can wait for the general release of >> the 3.7.x code. >> > >Ok.. I am requesting it!!! Please let me have it. Thank you... > Uh, I meant you have to "call" support.. I cant dist code. -m `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)
Subject: (usr-tc) new quad code
From: Brian <signal@shreve.net>
Date: 1998-02-04 10:24:03
Anyone put that new Quad code on there chassis yet? Does it help x2 any? Reading about it, its suppose to detect switch filters and stuff and adjust accordingly to improve x2 connections etc. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Netserver 8/I, radius, MPPP
From: Netlink Support Personnel <support@ns1.netlinkcom.com>
Date: 1998-02-04 10:43:03
What do I need to do to set up Multilink PPP on a Netserver 8/I? I am authenticating with radius and can set the ISDN port limit in radius to limit the user to 2 ports, but I am unsure of how to bond the 2 ports together with IP address. Thanks, Curt
Subject: Re: (usr-tc) new quad code
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-04 11:01:03
Brian said once upon a time: > >Anyone put that new Quad code on there chassis yet? Does it help x2 any? >Reading about it, its suppose to detect switch filters and stuff and >adjust accordingly to improve x2 connections etc. The bit about new Netserver code required made me weak in the knees. I'm waiting for that to be released before I begin the big upgrade.
Subject: Re: (usr-tc) Quake disconnects
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-04 11:31:37
At 10:22 AM 2/4/98 -0600, Mike wrote: >At 02:34 AM 2/4/98 -0500, you wrote: >> >> >>On Tue, 3 Feb 1998, Mike wrote: >> >>> At 03:19 PM 2/3/98 -0600, you wrote: >>> >At 01:51 PM 2/3/98 -0700, you wrote: >>> >>Brian Elfert said once upon a time: >>> >>> >>> >>>I have a couple of customers complaining that they get disconnected when >>> >>>playing Quake. >>> >>> >>> >>>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >>> >>>complained about any lag problems to date. >>> >> >>> >>Are they using X2? If so, reduce the top speed with &N. >>> >> >>> >You may also want to get the 3.6.69 Quake lag fix. >>> > >>> This is a fix available via support for Netserver & Quad Modem >>> combinations.. There are many improvements in the code path for >>> UDP traffic.. These improvements have also been pushed into the >>> 3.7.x code that is in BETA now.. >>> >>> We have mixed reports about improvement.. Some claim their lag went away, >>> some see improvement but not gone, and others claim that the improvement >>> is negligible.. If you want to try it, and are experiencing quake lag you >>> can request it through support or you can wait for the general release of >>> the 3.7.x code. >>> >> >>Ok.. I am requesting it!!! Please let me have it. Thank you... >> >Uh, I meant you have to "call" support.. I cant dist code. > Got it already... Thanks ;))) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: (usr-tc) double up cards
From: matthew <matthew@the-spa.com>
Date: 1998-02-04 12:27:02
what are people paying for a set of double up cards with the memory upgrades? best price i have so far is from source technology $12,298.00 matthew
Subject: Re: (usr-tc) double up cards
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-04 12:29:41
At 12:27 PM 2/4/98 -0500, matthew wrote: >what are people paying for a set of double up cards with the memory upgrades? > > best price i have so far is from source technology $12,298.00 > > > matthew Anyone have Source's Phone number? Apparently it changed from 617-890-4455 to something else.. And do they have a webpage? Can't seem to find that either.. Thanks for the info.. allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Michael H. Hamrich <mikeh@drfast.net>
Date: 1998-02-04 12:37:11
About the accounting logs not recording every event, and setting the Netserver to put out generic log files, Any ideas? I can't find anything in the documentation we have.
Subject: Re: (usr-tc) Merrit Radius
From: Michael H. Hamrich <mikeh@drfast.net>
Date: 1998-02-04 12:41:55
Yes, we would like only to have some help to point us to the source of this information. Not to do it for us., but were can you find out how to set the TC hub to only output group Radius accounting logs?
Subject: Re: (usr-tc) Quake disconnects
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-04 12:42:55
Brian said once upon a time: >its an improvement, not a fix, there is still bigtime latency issues even >with 3.6.69. > >Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. >Sometimes, you can get down to about 110-120. Brian, I'm curious as to whether your setup is running ISDN off the modems or of the Netserver.
Subject: Re: (usr-tc) X2 and Cellular options...
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-04 12:45:17
Brian Elfert said once upon a time: >On Tue, 3 Feb 1998, Pete Ashdown wrote: > >> >> Quad features not supported by Hiper DSP v1.0 >> >> --------------------------------------------- >> >> >> flow control (software OR hardware) >> > >> >How the heck do these modems even work without flow control? >> > >> >Won't they just keep sending data forever, even if the other end is >> >telling them to pause for a moment while the other end catches up? >> >> The packet bus handles it. The fact that Quads have it relates to the fact >> that some Quads have serial ports. > >Why mention the lack of things that don't even affect anything? I suppose they are either trying to cover their bases, or we're dealing with an overzealous technical writer. Some other features the HiPer lacks: Token ring Coffee Maker Video In/Out Fur
Subject: Re: (usr-tc) double up cards
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-04 13:02:19
On Wed, 4 Feb 1998, matthew wrote: > what are people paying for a set of double up cards with the memory upgrades? > > best price i have so far is from source technology $12,298.00 The last invoice I saw through here had 11,7xx. I don't remember the exact figure, and I don't know where they got them from either. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) Quake disconnects
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-04 14:51:47
Can you make Total Controls do ISDN DOV on T1 lines or only PRI? At 12:42 PM 2/4/98 -0700, Pete Ashdown wrote: >Brian said once upon a time: > >>its an improvement, not a fix, there is still bigtime latency issues even >>with 3.6.69. >> >>Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. >>Sometimes, you can get down to about 110-120. > >Brian, I'm curious as to whether your setup is running ISDN off the modems >or of the Netserver. > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) NETServer, routing, proxy-arp, subnet-0
From: Brian <signal@shreve.net>
Date: 1998-02-04 15:15:45
On Wed, 4 Feb 1998, Bob Purdon wrote: > > > > In addition, I need to have *some* dialups using addresses on the same > > > network as the net0 interface. Can I proxy-arp some dialups and route the > > > others? > > > > I have been there and it aint pretty :). If you *can*, try and get some > > more address space, and put your net0 interfaces on the same network as > > your normal LAN at your business. put the ptpx interfaces on there own > > network(s). > > That's what I was intending to do - with the dialup pools anyway. The > problem is that I've got several clients with static-IP's that are > currently in our LAN /24 space, and don't really want to move them, hence > I need to route most dialups, and proxy-arp a few. I understand, and when you proxy arp those people, using your LAN /24, your going to run into problems with arp caches. > > > We have a seperate network for static IP numbers than our dynamic pools. > > We route /26's to all of our hubs. > > Basically what I want to do, except the static IP's are on the same > network as net0. And this is going to bite you in the ass :). Move them, it only going to be worse later on when your 3 times as big, and have 3 times as many people eating at your lan address space and arp caches. > > > I ran into some serious problems of our Cisco adding an ARP entry for a > > user while they were dialed into TC-HUB1, and then the user hang's up, > > and dials into TC-HUB2 and can't get anywhere, because the Cisco's arp > > cache didn't expire. > > Not an issue for the moment. Our static-IP customers have dedicated modem > ports, so they'll always hit the same chassis. Ok, if you have dedicated modem ports, and they will *always* hit the same chassis, you are ok. > > > > I'm wanting to subnet a /24 into 4 * /26's and put the dialups in those. > > > Does the NETserver have any issues using the bottom and top subnets? If > > > so, I'll end up wasting half a /24 :-( > > > > No, the Netserver is CIDR aware. Unlike back in the days of "classfull" > > you *CAN* keep your upper and lower subnets today. > > Cool. Wasn't sure if the NETServer would do upper/lower or not - seems > like a 'hi-tech' feature for a box that only does RIP/RIPv2 :-) > Brian > Regards, > > Bob Purdon, > Technical Manager, > Southern Internet Services. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quake disconnects
From: Brian <signal@shreve.net>
Date: 1998-02-04 15:16:45
On Wed, 4 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > >its an improvement, not a fix, there is still bigtime latency issues even > >with 3.6.69. > > > >Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. > >Sometimes, you can get down to about 110-120. > > Brian, I'm curious as to whether your setup is running ISDN off the modems > or of the Netserver. ISDN off the modems, there were technical reasons for doing this, and I cant think of why right now. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quake disconnects
From: Charles Sprickman <spork@inch.com>
Date: 1998-02-04 15:21:18
Garry (and anyone else in the NYC area), I'm about to start tinkering with DOV very soon, and I will be dragging one of USR's local field technicians into this once I've got everything sorted out. If anyone would like to contribute any information, send me some email, and then I'll report back with full config info (USR & telco) when/if we make this work. It's just ridiculous that USR tech support seems unable to verify if it will or won't work, and even a definitive "no" would be nicer than all the messages we see on this list about the same topic. Lots of questions, but no answers... Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Wed, 4 Feb 1998, Garry Shtern wrote: > Date: Wed, 04 Feb 1998 14:51:47 -0500 > From: Garry Shtern <shterng@akula.com> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@lists.xmission.com, usr-tc@lists.xmission.com > Subject: Re: (usr-tc) Quake disconnects > > Can you make Total Controls do ISDN DOV on T1 lines or only PRI? > > At 12:42 PM 2/4/98 -0700, Pete Ashdown wrote: > >Brian said once upon a time: > > > >>its an improvement, not a fix, there is still bigtime latency issues even > >>with 3.6.69. > >> > >>Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. > >>Sometimes, you can get down to about 110-120. > > > >Brian, I'm curious as to whether your setup is running ISDN off the modems > >or of the Netserver. > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > > Garry Shtern shterng@akula.com > Chief Network Administrator http://www.akula.com > Akula Communications Corp. tel. (212) 292-8892 > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) double up cards
From: jcondon <jcondon@computer.net>
Date: 1998-02-04 15:24:50
Allen Marsalis wrote: > > Anyone have Source's Phone number? Apparently it changed from > 617-890-4455 to something else.. And do they have a webpage? Can't > seem to find that either.. Thanks for the info.. I have a number for Source Technology. Not sure if it is the same company you are looking for but, the do sell USR Equipment. 1-888-765-5758 Voice 1-888-867-6065 Fax
Subject: Re: (usr-tc) Quake disconnects
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-04 15:30:35
Please do tell.. If you need help, we'll work with you guys, since after all, we both have equipment at the same location... At 03:21 PM 2/4/98 -0500, Charles Sprickman wrote: >Garry (and anyone else in the NYC area), > >I'm about to start tinkering with DOV very soon, and I will be dragging >one of USR's local field technicians into this once I've got everything >sorted out. If anyone would like to contribute any information, send me >some email, and then I'll report back with full config info (USR & telco) >when/if we make this work. It's just ridiculous that USR tech support >seems unable to verify if it will or won't work, and even a definitive >"no" would be nicer than all the messages we see on this list about the >same topic. Lots of questions, but no answers... > >Charles > >~~~~~~~~~ ~~~~~~~~~~~ >Charles Sprickman Internet Channel >INCH System Administration Team (212)243-5200 >spork@inch.com access@inch.com > >On Wed, 4 Feb 1998, Garry Shtern wrote: > >> Date: Wed, 04 Feb 1998 14:51:47 -0500 >> From: Garry Shtern <shterng@akula.com> >> Reply-To: usr-tc@lists.xmission.com >> To: usr-tc@lists.xmission.com, usr-tc@lists.xmission.com >> Subject: Re: (usr-tc) Quake disconnects >> >> Can you make Total Controls do ISDN DOV on T1 lines or only PRI? >> >> At 12:42 PM 2/4/98 -0700, Pete Ashdown wrote: >> >Brian said once upon a time: >> > >> >>its an improvement, not a fix, there is still bigtime latency issues even >> >>with 3.6.69. >> >> >> >>Dual Channel 128k ISDN, will give you about a 130-150ping on 3.6.69. >> >>Sometimes, you can get down to about 110-120. >> > >> >Brian, I'm curious as to whether your setup is running ISDN off the modems >> >or of the Netserver. >> > >> >- >> > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> > with "unsubscribe usr-tc" in the body of the message. >> > For information on digests or retrieving files and old messages send >> > "help" to the same address. Do not use quotes in your message. >> > >> > >> >> Garry Shtern shterng@akula.com >> Chief Network Administrator http://www.akula.com >> Akula Communications Corp. tel. (212) 292-8892 >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the message. >> For information on digests or retrieving files and old messages send >> "help" to the same address. Do not use quotes in your message. >> > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) How to set Log group
From: steve_valiunas/mw/us/3com@usr.com
Date: 1998-02-04 15:39:21
--IMA.Boundary.652926688 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part The groups you refered to are used to control what the NMC sends to the accting server. They have no effect on what the netserver is reporting. "Mike Hamrich" <mhamrich@drfast.net> AT Internet on 01/28/98 02:00:09 PM cc: --IMA.Boundary.652926688 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part I have looked in all of the USR manuals. And found in the parametric reference page 3-15 that there are 8 group settings to controll the detail sent to the accouting servers. My questionion is how do you set the TC to use just group2 or what ever format that should work with Livingston Radius 2.01? Telneting to the netserver I fond these commands set accouting set accouting_options but they don't have enought help detail to show how to set it up Thanks - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. --IMA.Boundary.652926688 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4CF81240; Wed, 28 Jan 98 13:04:04 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id MAA18536; Wed, 28 Jan 1998 12:45:22 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0xxcdM-0004vi-00; Wed, 28 Jan 1998 11:54:48 -0700 Received: from drfast.net [209.57.81.2] by lists.xmission.com with smtp (Exim 1.73 #4) id 0xxcdJ-0004vJ-00; Wed, 28 Jan 1998 11:54:45 -0700 Received: from utbpopki [209.57.81.221] by drfast.net [209. 57.81.2] with SMTP (MDaemon.v2.6.rA.b6.32-R) for <usr-tc@lists.xmission.com>; Wed, 28 Jan 98 13:47:46 -0500 X-MSMail-Priority: Normal X-Priority: 3 X-Mailer: Microsoft Internet Mail 4.70.1161 MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-MDaemon-Deliver-To: usr-tc@lists.xmission.com Message-Id: <E0xxcdJ-0004vJ-00@lists.xmission.com> Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.652926688--
Subject: Re: (usr-tc) Quake disconnects
From: jcondon <jcondon@computer.net>
Date: 1998-02-04 15:45:30
Charles Sprickman wrote: > > Garry (and anyone else in the NYC area), > > I'm about to start tinkering with DOV very soon, and I will be dragging We are a small ISP in Westchester. We have a Netserver Imodem Plus 16 port. We too have heard it can and can not do DOV. We would be interested in what you find out. I currently have an account at another ISP just so I can do DOV (they are using Livingston portmasters). We had a few Zyxel Omni Ta's setup doing DOV but, they kept hanging every few days and the lines have been reused on the USR Netserver. Seems no one USR gives us the same answer. One said some about having it roll over on voice and/or data and it will work with some special command. Other techs had no idea or said no it will not work. One sales rep said to my boss it would work but, there is nothing in the manuals about 3.1Khz so I am skeptical. Unless it is an undoccumented command. I expected a better ping in Quake then we got from the Zyxel's (Ta's) pluged into a Computone Intelliserver. They seem a bit worse. Not alot worse but, not better. Figured no asyc to sync conversion. Guess these boxes have too much overhead or something. BTW, what USR server are you using?
Subject: RE: (usr-tc) List lag; 1200 bps connections
From: Kamil Kukura <k.kukura@unicom.cz>
Date: 1998-02-04 16:47:05
I noticed that when reflashing quads with newer code, it went to = setting &N to 1200bps. All it needs is setting them [quads] to factory = defaults. > -----P=F9vodn=ED zpr=E1va----- > Od: mark@vielle.datasys.net [SMTP:mark@vielle.datasys.net] > Odesl=E1no: 4. =FAnora 1998 3:03 > Komu: usr-tc@lists.xmission.com > P=F8edm=ECt: (usr-tc) List lag; 1200 bps connections >=20 . . . > Exactly one of those five racks performed normally when remounted and > fired > back up. I changed the call parameter settings, of course, and > expected > that to be my only change. But as it turned out, I've had to reflash > several > modems and fix a _lot_ of configuration information. In one case, ten > contiguous quad modems decided to take calls at 1200 bps -only-.=20 >=20 >=20
Subject: FW: (usr-tc) vendor specific attributes
From: Rudolph, Julie A. <julie.rudolph@teldta.com>
Date: 1998-02-04 17:12:41
i have contacted Merit regarding those messages in the logs and regarding the fact that, despite adding the appropriate entries to the dictionary, we have not been able to get Vendor-specific attributes (at least those from USR) to work. I have been told that there is a bug in their binary release of version 3.5, which is what we are running. A fix is supposedly due out soon. Please let me know if anyone discovers a workaround or something for this issue. We really need to have VSA's working properly. Julie. >------- >"Steve_Valiunas/MW/US/3Com@usr.com" AT Internet on 02/03/98 11:13:40 AM > >To: "Terry Kennedy" <terry@olypen.com> AT Internet@ccmail >cc: <usr-tc@lists.xmission.com> AT Internet@ccmail >Subject: Re: (usr-tc) vendor specific attributes > > > > > > > >You can find this in your dictnary.dat or documentation. The format is >four bytes of parameter followed by four bytes of >value, for example, Vendor-Specific=vUSR-00009023 0000000c = >parm x'9023' (connect speed) , value x'c' (21,600 bps). >You can control accounting info from the NMC's 'log group selection'. > > > > > >Using Merrit Radius these show up in our logs. Does anyone know how we can >identify these? I have looked in the reference I recieved with my TC's but >can't find a match. What is the format of the numbers at the end? Can you >control what values the TC's send to Radius? > >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-000090230000000c >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-0000006c00000013 >non-enapsulated vendor specific attribute >Vendor-Specific=vUSR-0000009900000003 >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-000000c700000001 >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-0000901900000000 >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-0000901a00000000 >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-0000901b00000000 >non-encapsulated vendor specific attribute >Vendor-Specific=vUSR-0000901d00000013 > > >Terry Kennedy Olypen, Inc. >terry@olypen.com >360 452 9755 > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with > "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages > send > "help" to the same address. Do not use quotes in your message. > > > > >Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP > (IMA Internet Exchange 2.02 Enterprise) id 4D0B0BA0; > Thu, 29 Jan 98 10:39:22 >-0600 >Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US > Robotics) > id KAA26733; Thu, 29 Jan 1998 10:20:41 -0600 (CST) >Received: from domo by lists.xmission.com with local (Exim 1.73 >#4) > id 0xxwuE-0002t5-00; Thu, 29 Jan 1998 09:33:34 -0700 >Received: from mail.olypen.com [208.200.248.2] > by lists.xmission.com with smtp (Exim 1.73 #4) > id 0xxwuB-0002sS-00; Thu, 29 Jan 1998 09:33:31 -0700 >Received: from modems.olypen.com(src addr [208.200.248.9]) (1462 >bytes) by >mail.olypen.com > via sendmail with P\:esmtp/R:inet_hosts/T:smtp > (sender: <terry@olypen.com>) > id <m0xxwuG-0002RjC@mail.olypen.com> > for <usr-tc@lists.xmission.com>; Thu, 29 Jan 1998 08:33:36 -0800 > (PST) > (Smail-3.2.0.96 1997-Jun-2 #19 built 1997-Jun-30) >Message-Id: <m0xxwuG-0002RjC@mail.olypen.com> >From: "Terry Kennedy" <terry@olypen.com> >To: <usr-tc@lists.xmission.com> >Subject: (usr-tc) vendor specific attributes >Date: Thu, 29 Jan 1998 08:34:10 -0800 >X-MSMail-Priority: Normal >X-Priority: 3 >X-Mailer: Microsoft Internet Mail 4.70.1161 >MIME-Version: 1.0 >Content-Type: text/plain; charset=ISO-8859-1 >Content-Transfer-Encoding: 7bit >Sender: owner-usr-tc@lists.xmission.com >Precedence: bulk >Reply-To: usr-tc@lists.xmission.com > > > > > > > >Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP > (IMA Internet Exchange 2.02 Enterprise) id 4D75A150; Tue, 3 Feb 98 > 11:55:33 >-0600 >Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) > id LAA03934; Tue, 3 Feb 1998 11:36:36 -0600 (CST) >Received: from domo by lists.xmission.com with local (Exim 1.73 #4) > id 0xzmT3-0007S4-00; Tue, 3 Feb 1998 10:49:05 -0700 >Received: from slack.xmission.com [199.104.120.18] (pashdown) > by lists.xmission.com with esmtp (Exim 1.73 #4) > id 0xzmSz-0007QD-00; Tue, 3 Feb 1998 10:49:01 -0700 >Received: (from pashdown@localhost) by slack.xmission.com (8.8.7/8.7.5) id >KAA13166 for usr-tc@lists.xmission.com; Tue, 3 Feb 1998 10:48:58 -0700 >(MST) >Received: from usr.com [149.112.20.2] > by lists.xmission.com with esmtp (Exim 1.73 #4) > id 0xzlxM-0004sw-00; Tue, 3 Feb 1998 10:16:20 -0700 >Received: from robogate2.usr.com by usr.com (8.8.5/3.1.090690-US Robotics) > id KAA02292; Tue, 3 Feb 1998 10:59:28 -0600 (CST) >Received: from ccMail by robogate2.usr.com > (IMA Internet Exchange 2.02 Enterprise) id 4D751510; Tue, 3 Feb 98 > 11:18:09 >-0600 >Mime-Version: 1.0 >Date: Tue, 3 Feb 1998 11:13:40 -0600 >Message-ID: <4D751510.3000@usr.com> >From: Steve_Valiunas/MW/US/3Com@usr.com >To: "Terry Kennedy" <terry@olypen.com> >Cc: <usr-tc@lists.xmission.com> >Subject: Re: (usr-tc) vendor specific attributes >Content-Type: multipart/mixed; boundary="IMA.Boundary.882625688" >Sender: owner-usr-tc@lists.xmission.com >Precedence: bulk >Reply-To: usr-tc@lists.xmission.com > > > > > > >
Subject: Re: (usr-tc) double up cards
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-04 21:22:47
At 03:24 PM 2/4/98 -0500, jcondon wrote: >Allen Marsalis wrote: >> >> Anyone have Source's Phone number? Apparently it changed from >> 617-890-4455 to something else.. And do they have a webpage? Can't >> seem to find that either.. Thanks for the info.. > >I have a number for Source Technology. Not sure if it is the same >company you are looking for but, the do sell USR Equipment. > >1-888-765-5758 Voice >1-888-867-6065 Fax > yes it is.. Thanks. I also found their direct numbers: voice 781-893-5530 fax 781-893-7220 and I think their webpage is www.source-technology.com. I hate those dashes in domain names.. allen
Subject: (usr-tc) Hyper about HiPer
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-04 21:49:13
A while back I related our experience with HiPer DSPs in our hunt group; we had several loop-start channelized T1s and 32 1FB (regular analog lines) after all of the T1s. It appeared that as soon as the T1s filled up and a call rolled over to the 1FBs, the HDM would drop all of its connections. The people at USR/3Com have been very nice and, it appears, have put significant work into this issue. We moved our entire operation recently, and had the channelized T1s provisioned as ground-start at the new site. Yesterday, Oscar from 3Com installed some new code on our HDM, et voila, it appears to be functioning perfectly. I hate the fact that we changed two variables at once and have only one result (wouldn't that be an undetermined linear system?), but that result is a good one. Of course, we're happy, but we're keeping an eye on it to be sure it doesn't flake out again.
Subject: Re: (usr-tc) NETServer, routing, proxy-arp, subnet-0
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-04 21:59:59
> > In addition, I need to have *some* dialups using addresses on the same > > network as the net0 interface. Can I proxy-arp some dialups and route the > > others? > > I have been there and it aint pretty :). If you *can*, try and get some > more address space, and put your net0 interfaces on the same network as > your normal LAN at your business. put the ptpx interfaces on there own > network(s). That's what I was intending to do - with the dialup pools anyway. The problem is that I've got several clients with static-IP's that are currently in our LAN /24 space, and don't really want to move them, hence I need to route most dialups, and proxy-arp a few. > We have a seperate network for static IP numbers than our dynamic pools. > We route /26's to all of our hubs. Basically what I want to do, except the static IP's are on the same network as net0. > I ran into some serious problems of our Cisco adding an ARP entry for a > user while they were dialed into TC-HUB1, and then the user hang's up, > and dials into TC-HUB2 and can't get anywhere, because the Cisco's arp > cache didn't expire. Not an issue for the moment. Our static-IP customers have dedicated modem ports, so they'll always hit the same chassis. > > I'm wanting to subnet a /24 into 4 * /26's and put the dialups in those. > > Does the NETserver have any issues using the bottom and top subnets? If > > so, I'll end up wasting half a /24 :-( > > No, the Netserver is CIDR aware. Unlike back in the days of "classfull" > you *CAN* keep your upper and lower subnets today. Cool. Wasn't sure if the NETServer would do upper/lower or not - seems like a 'hi-tech' feature for a box that only does RIP/RIPv2 :-) Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: Re: (usr-tc) Hyper about HiPer
From: Brian <signal@shreve.net>
Date: 1998-02-04 22:13:00
On Wed, 4 Feb 1998, Mark R. Lindsey wrote: > A while back I related our experience with HiPer DSPs in our hunt group; > we had several loop-start channelized T1s and 32 1FB (regular analog > lines) after all of the T1s. It appeared that as soon as the T1s filled > up and a call rolled over to the 1FBs, the HDM would drop all of its > connections. > > The people at USR/3Com have been very nice and, it appears, have put > significant work into this issue. We moved our entire operation > recently, and had the channelized T1s provisioned as ground-start at the > new site. > > Yesterday, Oscar from 3Com installed some new code on our HDM, et voila, > it appears to be functioning perfectly. I hate the fact that we changed > two variables at once and have only one result (wouldn't that be an > undetermined linear system?), but that result is a good one. > > Of course, we're happy, but we're keeping an eye on it to be sure it > doesn't flake out again. I have noticed on one of our HiPerDSP's, which has a PRI in it, that it will just all the suddent dump all calls, and refuse any new calls until its been reset. I only saw this happen one time, but am suspicious. Is this a known problem? Brian > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Java for USRTC
From: Craig Nelson <craig@jetcity.com>
Date: 1998-02-04 22:38:44
I have created a prototype JAVA TC Chassis monitor. It is still pretty ugly, but it uses both SNMP and a telnet screen-scrape process to continuosly track the status of all the active ports on any number of chassis (I have three- but it should scale much higher). Right now it runs on my Windows machine, but the plan is to run it as a daemon on my Solaris box (not that the OS matters) and have it just run continuously. It has a very simple web server built into it right now, and users can then connect to the daemon to get their current connection status, SNMP variables, connect time, packects in and out, all the stuff. Later another mode could be added for more active management, like resetting ports, kicking simultaneous users off, all that. Brent Jay wrote: > > On Mon, 2 Feb 1998, Brian wrote: > > > I think that Java is "ideal" for applications such as TCM. There are lots > > of programs, such as TCM, where scalability demands outweigh perforamnce. > > The performance of Java isn't that bad, and the "write once" idea I really > > like. > > > > Anyone know if USR sees the light and is thinking of making Java managment > > applications for its Total Control gear? > > > > I also think some products, could benefit by having small Webservers built > > into them (kinda like cisco), where you just http to there IP, and bring > > up a managment frontend, in Java of course. > > There are web servers built into a few 3Com products. The ADSL products > like the Viper or Axcell card both have it and the Lan-linkers as well. > It is basically just netstats and routes and information about the unit. > It is basic, but it is a start. (: > > :::::::::::::::::::::::::::::::::::: > :: :: > :: bjay@ionet.net :: > :: ioNET network specialist :: > :: break out the blender and :: > :: mix me a spam margarita! :: > :: 1-800-360-5183 405-270-0999 :: > :: :: > :::::::::::::::::::::::::::::::::::: > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- <.>-----------------------------------------------<.> The thing about Crayons is, They can take you more places than Starships -Guinan <.>-----------------------------------------------<.>
Subject: Re: (usr-tc) Hyper about HiPer
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-05 00:31:34
Brian said: : I have noticed on one of our HiPerDSP's, which has a PRI in it, that it : will just all the suddent dump all calls, and refuse any new calls until : its been reset. I only saw this happen one time, but am suspicious. This was the behavior we observed orginally. The 3Com people made some initial changes that made it start accepting calls immediately after it had dumped the formerly-happy surfers. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: (usr-tc) HiPer ReBoot
From: Brian <signal@shreve.net>
Date: 1998-02-05 08:58:43
Anyone noticing there ARC rebooting at random times? I am opening a ticket on this now, just wanted to post here to see whats up with that. code version 4.0.19. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NETServer, routing, proxy-arp, subnet-0
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-05 09:00:25
> > Basically what I want to do, except the static IP's are on the same > > network as net0. > > And this is going to bite you in the ass :). Move them, it only going to > be worse later on when your 3 times as big, and have 3 times as many > people eating at your lan address space and arp caches. Yeah, I know. For various reasons I can't move the static IP's just yet, but do need to move the dialup pools fairly soon. The original design of this network (before my time) had all dialup IP's in the LAN space and it's come time to move them because I need the address space back. > > Not an issue for the moment. Our static-IP customers have dedicated modem > > ports, so they'll always hit the same chassis. > > Ok, if you have dedicated modem ports, and they will *always* hit the same > chassis, you are ok. Yup. It will become an issue in the future, but by then I'll be able to relocate the static IP customers. Thanks for confirming that what I want to do will work (albeit with some limitations) :-) Regards, Bob.
Subject: Re: (usr-tc) HiPer ReBoot
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-05 09:10:20
> Anyone noticing there ARC rebooting at random times? I am opening a > ticket on this now, just wanted to post here to see whats up with that. > code version 4.0.19. The ARC code is basic right now... Try and do too much, and you are going to have reboots. It also seems VERY unstable to tinkering. My suggestion is to busy out the box and tinker, then let a few people on, and watch it for at least an hour. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: (usr-tc) Multiple Netservers???
From: Terry Kennedy <terry@olypen.com>
Date: 1998-02-05 09:21:05
Following the post on multiple netservers... What's the point? Are the netservers incapable of handling 48 X2 modems calls? Is there a performance boost here?
Subject: (usr-tc) Two Netservers in one chassis
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-05 10:00:35
I just want to verify that I can put two Netservers in one chassis, and have them each handle some of my quad modems. I assume I'd just set some some modems active in each Netserver, and away I'd go. What I want to do is have each Netserver point to a different Radius server. Brian
Subject: (usr-tc) For Sale: Used MP16/I-Modem
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-05 10:43:48
For Sale: TOTAL CONTROL MP16/I MODEM ISDN & ANALOG MODEM POOL 16-LINES Has been in use for 3 months w/o any problems and currently in perfect working condition. X2 code has been enabled, was plugged into Portmaster-2E-30 and 8 BRI lines to offer X2 connections. Can service both X2 and ISDN connections. Asking $3700. If interested or need more info, please email back privately. Thanks.
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Mike Wronski <mwronski@coredump.ae.usr.com>
Date: 1998-02-05 11:06:03
On Thu, 5 Feb 1998, Brian Elfert wrote: > I just want to verify that I can put two Netservers in one chassis, and > have them each handle some of my quad modems. > > I assume I'd just set some some modems active in each Netserver, and away > I'd go. What I want to do is have each Netserver point to a different > Radius server. > This will work fine... You can even make each netserver start its first modem at S1 using the 'set modem startslot.." command. Make sure that there is no conflict on who owns which modem.. -m
Subject: Re: (usr-tc) HiPer ReBoot
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-05 11:26:44
Brian said once upon a time: > >Anyone noticing there ARC rebooting at random times? I am opening a >ticket on this now, just wanted to post here to see whats up with that. >code version 4.0.19. Yes, I see it on 4.0.89. The last technician I talked to about this stated that I should reflash using ZModem. That hasn't help. Looks like I've got to put a logging console on the ARC again.
Subject: Re: (usr-tc) hiperarc question?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-05 11:27:49
MyungSik Kim said once upon a time: > > > I have several question on hiperarc. Is it support rfc1570, rlogin -l >(for connecting rlogin host directly without any prompt, like autolog in >netserver) and d-channel up/down trap? It supports "rlogin -l".
Subject: (usr-tc) Total Control Admin-Reset Bug
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-05 11:35:50
When manually resetting ports on the Total Control (admin-reset), the account session time is reported as zero in the radius accounting logs (we are running Livingston radius 1.16). Resetting ports from our Portmasters do not result in this bug. Has anyone else reported this bug? Is this bug confirmed by 3COM/USR? We usually manually reset ports for customers that hog the line for 1-2 days, and if it doesn't report that time, then that is 1-2 days worth of login time lost to our logs! Sun Feb 1 16:53:30 1998 Acct-Session-Id = "0602057A" User-Name = "bowman" Client-Id = 205.131.158.51 Client-Port-Id = 27 Acct-Status-Type = Stop Acct-Session-Time = 0 Acct-Terminate-Cause = Admin-Reset Acct-Authentic = RADIUS Called-Station-Id = "6331" Acct-Input-Octets = 734511 Acct-Output-Octets = 889523 NAS-Port-Type = Async User-Service-Type = 2 Framed-Protocol = PPP Framed-Address = 205.131.158.45 Acct-Delay-Time = 0 | --------------------------------------------------- | Lee Kuo | Take back the web! Use a third party browser: | | lee@mitec.net | Fast - http://www.operasoftware.com | ---------------- Free - http://www.mnemonic.org |
Subject: (usr-tc) Callback between a PC/Linux and an USR Total Control
From: Stephane Bortzmeyer <bortzmeyer@pasteur.fr>
Date: 1998-02-05 11:38:38
I'm trying to call an USR Total Control from a PC/Linux and to get the TC call me back. I control both sides, so I can change what I wish. I know how to do it in POTS (define a "location" on the Netserver of the TC, define an user - Radius or local - which will call through the above location). I would like to do it in ISDN (and, if I can get it to work without the PC owner paying the communication AT ALL, it would be even better -it seems possible with SIDN). I didn't find a simple HOWTO. Any idea? PC/Linux Debian 1.3, kernel 2.0.29 with the Teles 16.3 card (I can upgrade the kernel or isdn4utils if necessary).
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-05 11:46:02
On Thu, 5 Feb 1998, Mike Wronski wrote: > > I'd go. What I want to do is have each Netserver point to a different > > Radius server. > > > This will work fine... You can even make each netserver start its first > modem at S1 using the 'set modem startslot.." command. Make sure that > there is no conflict on who owns which modem.. Does "set modem startslot" work in 3.6.28? This command first appeared in 3.4.23 or so, but I don't think it worked right in 3.4.23. I tried it with 3.4.23 and hosed up the entire Netserver. I called support, and they told me to clear the config, and not to use the command. Brian
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-05 11:46:03
At 11:06 AM 2/5/98 -0600, Mike Wronski wrote: > > >On Thu, 5 Feb 1998, Brian Elfert wrote: > >> I just want to verify that I can put two Netservers in one chassis, and >> have them each handle some of my quad modems. >> >> I assume I'd just set some some modems active in each Netserver, and away >> I'd go. What I want to do is have each Netserver point to a different >> Radius server. >> >This will work fine... You can even make each netserver start its first >modem at S1 using the 'set modem startslot.." command. Make sure that >there is no conflict on who owns which modem.. > Might this have been a good quake lag solution??.. I am surprised that someone didn't suggest lightening the load on the netserver by limiting it's scope to one span only.. (two per chassis) With the recent code improvements to boot, I bet quake lag could all but be eliminated nowadays without a hiperarc.. Which still has *some* lag in my opinion.. (Looks like I'll be trying two arcs.. :) I'm not knocking the arc or anything. It's great if you really need the high densitys. And we did have to install custom A/C which runs year around. Without hdms, a big problem would have slowly developed once again. And overall ping (both quake and ICMP) times are better. But still not as good as a trusty ole 486 with a sportster.. (talk about low densitys..) The reason I mention this is we are moving our 2059's to remote pops to hasten hdm/arc upgrades here at home.. And these pops are of course another hop away (from our quake server).. With used netservers coming down in price, I am interested in an "affordable" solution to quake lag. We just can't put hiperarcs (maybe two?) in all our chassis without some help from usr in the form of an upgrade path.. A used netserver, i could probably swing... Any thoughts? allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Multiple Netservers???
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-05 11:52:37
On Thu, 5 Feb 1998, Terry Kennedy wrote: > Following the post on multiple netservers... What's the point? > Are the netservers incapable of handling 48 X2 modems calls? > Is there a performance boost here? A local ISP had some kind of problem, and lost their phone service for several weeks. They approached us about providing dialup service for a few weeks. We have a lot of extra capacity, and we could get the telco to pull one of our CT1s out of our main hunt group for a few weeks. I have an spare Netserver card, so I want to use that for the other ISP's stuff, so they can use their radius server and DNS server. 24 of the modems in the rack will be ours, and 24 of the modems will be for the other ISP. Brian
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-05 13:32:07
> Does "set modem startslot" work in 3.6.28? > > This command first appeared in 3.4.23 or so, but I don't think it worked > right in 3.4.23. I tried it with 3.4.23 and hosed up the entire > Netserver. I called support, and they told me to clear the config, and > not to use the command. Yes, it works great. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-05 13:34:17
> > I'm not knocking the arc or anything. It's great if you really need > the high densitys. And we did have to install custom A/C which runs > year around. Without hdms, a big problem would have slowly developed Tell me about it! We've had to run cooling year round for years. We did manage to get a decent economizer solution working, though, which sucks outside cold air in the winter so that you don't end up using a compressor of an A/C unit. Our current modem room has a 12 ton A/C unit all to itself... <ouch> | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) HiPer ReBoot
From: Michael Mittelstadt <meek@execpc.com>
Date: 1998-02-05 14:28:16
[Quoth Pete Ashdown] ] Brian said once upon a time: ] > ] >Anyone noticing there ARC rebooting at random times? I am opening a ] >ticket on this now, just wanted to post here to see whats up with that. ] >code version 4.0.19. ] ] Yes, I see it on 4.0.89. The last technician I talked to about this stated ] that I should reflash using ZModem. That hasn't help. Looks like I've got ] to put a logging console on the ARC again. Do you have service_loss_busyout set to radius? If I do that, I see crashes. Do you have any packet filters? If I enable any packet filters on the modem interfaces, I also see crashes. And, as Curt said, if we tinker with things for any length of time, the ARC gets frustrated with us, and crashes. On another note. Does the "set syslog ____ loglevel" command work right for anyone? I set it to 'CRITICAL', and I'm still seeing hundreds of 'COMMON' messages being logged, even after a reboot. -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1
Subject: Re: (usr-tc) HiPer ReBoot
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-05 15:08:22
Michael Mittelstadt said once upon a time: >Do you have service_loss_busyout set to radius? If I do that, I see >crashes. Do you have any packet filters? If I enable any packet >filters on the modem interfaces, I also see crashes. And, as Curt >said, if we tinker with things for any length of time, the ARC gets >frustrated with us, and crashes. I have neither service_loss or filters set, and it crashes. >On another note. Does the "set syslog ____ loglevel" command work >right for anyone? I set it to 'CRITICAL', and I'm still seeing >hundreds of 'COMMON' messages being logged, even after a reboot. This is a bug in 4.0.19.
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-05 15:11:38
> When will Livingston post a comparison of yearly savings in A/C and power > cunsumption costs of the PM3 vs. the TCH? ;) > > Would you mind sharing what your yearly costs are to run a 12 ton A/C > unit? No clue actually - it's on the same meter as the rest of our 480v 3-phase A/C units. I do know I needed, recently, to upgrade our main feed to the building (the single phase) from 400AMP to 800AMP to keep up with our expansions. Not all of the load is total control, of course, but I have a good 8 20-amp circuits dedicated to total controls. I've actually been only mildly dissapointed with the total control power consumption. With stand alone modems, I could get about 120 modems running for the same power I can now get 96 running with total control. Not too bad considering the advantages, and the fan tray. Also, with our new HiPer box with 336 ports and 2 ARC cards - it only takes slightly more than our old Total control boxes with 48 ports. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Charles Sprickman <spork@inch.com>
Date: 1998-02-05 15:33:59
When will Livingston post a comparison of yearly savings in A/C and power cunsumption costs of the PM3 vs. the TCH? ;) Would you mind sharing what your yearly costs are to run a 12 ton A/C unit? Charles > > Tell me about it! We've had to run cooling year round for years. We did > manage to get a decent economizer solution working, though, which sucks > outside cold air in the winter so that you don't end up using a compressor > of an A/C unit. Our current modem room has a 12 ton A/C unit all to > itself... <ouch> > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) newbie: rlogin via NETServer/16
From: FBSD <fbsd@typhoon.co.jp>
Date: 1998-02-05 17:40:21
Hello, I've just purchased a NETServer/16 I-modem and I'm having problems with rlogin. When a "login-user" dials in (say mod:n) and rlogin to a host, the session would hang after a few seconds, or after a few "ls -la" If another "login-user" dials in again to the same interface, it would hang after the "login: " prompt. Syslog shows the following: ---syslog extracts starts--- Feb 5 17:13:09 netserver At 17:14:21, Facility "Telnet", Level "UNUSUAL":: Overflow on escape queue. Queue size : 0 Receive size : 14 Feb 5 17:13:27 netserver At 17:14:39, Facility "Telnet", Level "UNUSUAL":: Bad state/event combination: link status update received in state: 0x9 Feb 5 17:14:00 netserver At 17:15:12, Facility "Telnet", Level "UNUSUAL":: Received unknown message : 0x80565244 Feb 5 17:14:13 netserver At 17:15:25, Facility "Telnet", Level "UNUSUAL":: Received unknown message : 0x84565244 Feb 5 17:14:30 netserver At 17:15:43, Facility "Telnet", Level "UNUSUAL":: Bad state/event combination: link status update received in state: 0x9 ---syslog extracts ends--- Login-users using telnet have no problems... System information: ---SYSTEM DESCRIPTION--- System Descriptor: U.S. Robotics NetServer/16I, Version: V4.0.14, Built on Sep 11 1997 at 13:10:08. Object ID: 1.3.6.1.4.1.429.2.15 System UpTime: 0d 00:28:56 System Services: Internet EndToEnd Applications System Version: V4.0.14
Subject: RE: (usr-tc) hdm+netserver
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-05 17:58:48
With E1 PRI, we had a nice problem using an old netserver with 3.6.28, the cards were overloaded at 40-50 calls. The Netserver accepted the call and *POP* just disconnects the user. I have no limits set on the netserver, and it sees both cards (set to round-robin, S1-S60 get used). I have 16 meg in the card. I replaced the card with a HiPerARC, and *whoosh*, no problems, 60 concurrent calls, users lots happier... I have a ticket open with USR Europe, but they are very surprised, checked my Hub, found no problems. Very strange, it's not a telco problem, and not a configuration problem with the HDM's (didn't change the config), so it has to be with the NetServer. Any ideas ? Robert von Bismarck Petrel Communication S.A. -----Original Message----- From: Tatai SV Krishnan [SMTP:tkrishna@bubba.ae.usr.com] Sent: vendredi, 23. janvier 1998 02:37 To: Brian Cc: USRobotics TC Mailing List Subject: Re: (usr-tc) hdm+netserver Four Total of 96 ports krish On Thu, 22 Jan 1998, Brian wrote: > How many HDM's can you plug into a chassis with an EPB netserver? > > I am talking about the newer chassis, high-speed backplane, packet bus > clocking. > > Brian > > > /-------------------------- signal@shreve.net > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Systems Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.3 | > \-------------------------- 318-222-2638 x109 > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > ----------------------------------------- \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ -\ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec -/ - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: (usr-tc) CPU Load on HiPerARC by SNMP ??
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-05 18:05:37
The title says it all... On my Cisco 3640 and 7206 I can SNMP-get the actual CPU load, can I do the same on a HiPerARC board=A0? Any info will be appreciated, because I'm trying to get about = everything useful onto my MRTG pages (like temp, capacity, etc...) Robert von Bismarck Petrel Communication S.A.
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Brian <signal@shreve.net>
Date: 1998-02-05 20:37:46
On Thu, 5 Feb 1998, Brian Elfert wrote: > I just want to verify that I can put two Netservers in one chassis, and > have them each handle some of my quad modems. > > I assume I'd just set some some modems active in each Netserver, and away > I'd go. What I want to do is have each Netserver point to a different > Radius server. I think I am going to do this. After you upgrade a chassis with a ARC you got a left over Netserver, which is great for more cpu power in a chassis. Probably would reduce quake lag as well. Brian > > Brian > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) CPU Load on HiPerARC by SNMP ??
From: Brian <signal@shreve.net>
Date: 1998-02-05 20:39:23
On Thu, 5 Feb 1998, Robert von Bismarck wrote: > The title says it all... >=20 > On my Cisco 3640 and 7206 I can SNMP-get the actual CPU load, can I do > the same on a HiPerARC board=A0? >=20 > Any info will be appreciated, because I'm trying to get about everything > useful onto my MRTG pages (like temp, capacity, etc...) >=20 > Robert von Bismarck > Petrel Communication S.A. I really like the snmp/mrtg stuff, perhaps when you get it all pounded out you can share some of your ideas with us, or let us check your web page out. Brian >=20 >=20 > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >=20 /-------------------------- signal@shreve.net -----------------------------= \ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 = | | Network Administrator | Perl, Linux | Web hosting, online stores, = | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs = | | 89 CRX DX w/MPFI, lots of |-=3D*:Quake:*=3D-| http://www.shreve.net/ = | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 = | \-------------------------- 318-222-2638 x109 -----------------------------= /
Subject: Re: (usr-tc) HiPer ReBoot
From: Brian <signal@shreve.net>
Date: 1998-02-05 20:44:03
On Thu, 5 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > > >Anyone noticing there ARC rebooting at random times? I am opening a > >ticket on this now, just wanted to post here to see whats up with that. > >code version 4.0.19. > > Yes, I see it on 4.0.89. The last technician I talked to about this stated > that I should reflash using ZModem. That hasn't help. Looks like I've got > to put a logging console on the ARC again. pete, I opened up a ticket on this #38888, get 4.0.80, its the fix for 4.0.89 rebooting. I sent 3com my "crashdump" and from that they said I needed 4.0.80 and that should fix it. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) HiPer ReBoot
From: Brian <signal@shreve.net>
Date: 1998-02-05 20:49:09
On Thu, 5 Feb 1998, Michael Mittelstadt wrote: > [Quoth Pete Ashdown] > ] Brian said once upon a time: > ] > > ] >Anyone noticing there ARC rebooting at random times? I am opening a > ] >ticket on this now, just wanted to post here to see whats up with that. > ] >code version 4.0.19. > ] > ] Yes, I see it on 4.0.89. The last technician I talked to about this stated > ] that I should reflash using ZModem. That hasn't help. Looks like I've got > ] to put a logging console on the ARC again. > > Do you have service_loss_busyout set to radius? If I do that, I see > crashes. Do you have any packet filters? If I enable any packet radius busyout is not set, we have 2 seperate very reliable radius servers. I do have packet filters defined on the netservers, but not on the ARC. I wonder, though, are you saying that a user who has a filter assigned in radius such as "mail.in" could crash my arc because the packet filter is not defined on the arc? > filters on the modem interfaces, I also see crashes. And, as Curt > said, if we tinker with things for any length of time, the ARC gets > frustrated with us, and crashes. > I don't tinker at all. I set the initial "quick setup", defined a ip pool, and thats really it. > On another note. Does the "set syslog ____ loglevel" command work > right for anyone? I set it to 'CRITICAL', and I'm still seeing > hundreds of 'COMMON' messages being logged, even after a reboot. > > -- > Michael Mittelstadt meek@execpc.com > VP - Internet Techologies ExecPC Internet > http://www.execpc.com/~meek 1-800-ExecPC-1 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Total Control Admin-Reset Bug
From: Brett Hawn <blh@texas.net>
Date: 1998-02-05 21:21:15
Upgrade your Netserver/NMC code, we had this problem as well and it was fixed a version or two ago. On Thu, Feb 05, 1998 at 11:35:50AM -0600, Lee Kuo stated > > When manually resetting ports on the Total Control (admin-reset), the > account session time is reported as zero in the radius accounting logs (we > are running Livingston radius 1.16). Resetting ports from our Portmasters > do not result in this bug. Has anyone else reported this bug? Is this > bug confirmed by 3COM/USR? We usually manually reset ports for customers > that hog the line for 1-2 days, and if it doesn't report that time, then > that is 1-2 days worth of login time lost to our logs! > > Sun Feb 1 16:53:30 1998 > Acct-Session-Id = "0602057A" > User-Name = "bowman" > Client-Id = 205.131.158.51 > Client-Port-Id = 27 > Acct-Status-Type = Stop > Acct-Session-Time = 0 > Acct-Terminate-Cause = Admin-Reset > Acct-Authentic = RADIUS > Called-Station-Id = "6331" > Acct-Input-Octets = 734511 > Acct-Output-Octets = 889523 > NAS-Port-Type = Async > User-Service-Type = 2 > Framed-Protocol = PPP > Framed-Address = 205.131.158.45 > Acct-Delay-Time = 0 > > | --------------------------------------------------- > | Lee Kuo | Take back the web! Use a third party browser: | > | lee@mitec.net | Fast - http://www.operasoftware.com | > ---------------- Free - http://www.mnemonic.org | > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- <HTML><META HTTP-EQUIV="Content-Type:text/html"> <SCRIPT> function X() {var Text = "HTML is not acceptable for using in mail " + "or usenet so your browser will stop."; alert(Text); parent.close();}; </SCRIPT> </HEAD><BODY onLoad="X();return true">Hi</HTML>
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-05 21:56:45
At 01:34 PM 2/5/98 -0600, Curt Shambeau wrote: >Tell me about it! We've had to run cooling year round for years. We did >manage to get a decent economizer solution working, though, which sucks >outside cold air in the winter so that you don't end up using a compressor >of an A/C unit. Our current modem room has a 12 ton A/C unit all to >itself... <ouch> We installed a 5 ton unit after our landlord quoted us $8/hr for overtime heating and air.. I calculated that to be $3500/mo overtime on a $1500/mo office! Now we just pay for utility costs to operate the unit. We have 1500 sq feet and our modem room is the return air "plentum". There is enough heat generated by our gear to heat the entire office and call on the a/c compressor at times. Here in the south, we only get a few hard freezes per winter but I'm waiting for our compressor to freeze up one cold wet winter day.. We'll probably be ok this summer but we are growing rapidly. Without hdms, I would have to purchase HVAC equipment in addition to my current $15K investment by summer '99 for sure. But now I figure I can handle 1000 to 1500 modems without additional a/c.. These kind of expenses/savings should be considered in business projections. We did not count on (nor budget) any a/c expenses originally. Coming up with $15K was a big surprise. allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: (usr-tc) hiperarc question?
From: MyungSik Kim <daesung@soback.kornet.nm.kr>
Date: 1998-02-05 22:43:53
I have several question on hiperarc. Is it support rfc1570, rlogin -l (for connecting rlogin host directly without any prompt, like autolog in netserver) and d-channel up/down trap?
Subject: Re: (usr-tc) HiPer ReBoot
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-06 02:33:17
On Fri, 6 Feb 1998, Curt Shambeau wrote: > > 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they number > > their engineering releases. > > Their release versions go up, their engineering releases start at the end > of group (xx.xx.99) and go down. I wonder what happens if they meet in > the middle?? <grin> The way the ER release works is that we start at end xx.xx.99 and go down. It should not meet in the middle, there are a certain number of ER releases before which we will have a new release with a different number. krish > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) ARC/HDM problem
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-06 08:32:48
make sure that you have the chassis awarness enabled - if you have only one HiPer ARC. ena nmc cha If you have more than one HiPer ARC then you have to set it static - at this point - will be fixed soon. krish On Fri, 6 Feb 1998, Brian wrote: > I added some hdm's today, and the # of "ports" on them does not look > right. Both are configured for PRI, and should show 23 ports: > > HiPer>> list chassis > Slot Owner Description Ports Type > . > . > . > 10 YES 24 Channel High Density Modem 23 DYNAMIC > 11 YES 24 Channel High Density Modem 23 DYNAMIC > 12 YES 24 Channel High Density Modem 23 DYNAMIC > 13 YES 24 Channel High Density Modem 23 DYNAMIC > 14 YES 24 Channel High Density Modem 24 DYNAMIC > 15 YES 24 Channel High Density Modem 0 DYNAMIC > > as you can see, slot 14 shows 25 and slot 15 shows 0! At least slot 14 > will take calls, slot 15 has 0 ports so will just return busies. > > I *cant* statically set the # of ports, because I have an nmc card in this > hub, and its suppose to handle this. All you can really set statically on > an arc is card_type and ownership. > > Anyways, here is what a partial list interfaces shows: > . > . > . > slot:14/mod:22 Up Up > slot:14/mod:23 Up Up > slot:14/mod:24 Down Down > internal Up Up > loopback Up Up > > > as you can see slot:14 goes to 24......even though it *is* marked "down" I > don't like it, because my other HDM's only show 23. Also you can see > there is a total absense of slot:15! > > I have tried to re-insert the card to no avail. > > If this were a netserver, I could fix it. its almost like I need to do a > "set density" or a "set modem active" but I don't know how to do that on a > ARC! > > Any help would be greatly appreciated. > > Brian > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) Two Netservers in one chassis
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-06 09:26:06
> We'll probably be ok this summer but we are growing rapidly. Without hdms, > I would have to purchase HVAC equipment in addition to my current $15K > investment by summer '99 for sure. But now I figure I can handle 1000 to > 1500 modems without additional a/c.. These kind of expenses/savings should > be considered in business projections. We did not count on (nor budget) > any a/c expenses originally. Coming up with $15K was a big surprise. I would say your predictions are probably accurate. We had 5 tons feeding our modem room, and it started getting totally clobbered at around 2000 modems - to the point of an 80 degree modem room. Now, with almost 4000 ports, plus around 30 computers, it's a healthy 69 degrees all year round. I think we ended up spending around $35k for our 12 ton unit, which I felt wasn't too bad for how well it performs. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) HiPer ReBoot
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-06 11:03:07
Brian said once upon a time: >> Yes, I see it on 4.0.89. The last technician I talked to about this stated >> that I should reflash using ZModem. That hasn't help. Looks like I've got >> to put a logging console on the ARC again. > >pete, I opened up a ticket on this #38888, get 4.0.80, its the fix for >4.0.89 rebooting. I sent 3com my "crashdump" and from that they said I >needed 4.0.80 and that should fix it. 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they number their engineering releases.
Subject: Re: (usr-tc) HiPer ReBoot
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-06 12:12:16
> 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they number > their engineering releases. Their release versions go up, their engineering releases start at the end of group (xx.xx.99) and go down. I wonder what happens if they meet in the middle?? <grin> | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) HiPer ReBoot
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-06 12:18:06
On Fri, 6 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > >> Yes, I see it on 4.0.89. The last technician I talked to about this stated > >> that I should reflash using ZModem. That hasn't help. Looks like I've got > >> to put a logging console on the ARC again. > > > >pete, I opened up a ticket on this #38888, get 4.0.80, its the fix for > >4.0.89 rebooting. I sent 3com my "crashdump" and from that they said I > >needed 4.0.80 and that should fix it. > > 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they number > their engineering releases. > Maybe he ment 4.0.90? :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) HiPer ReBoot
From: Charles Sprickman <spork@inch.com>
Date: 1998-02-06 13:49:16
This is a scary look into the collective mind of 3com engineering... There had to have been more than one person that said "yeah, that's a *great* idea!". > Their release versions go up, their engineering releases start at the end > of group (xx.xx.99) and go down. I wonder what happens if they meet in > the middle?? <grin> > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) Open code
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-06 14:26:00
I've been discussing this idea with a friend, and although it may never fly its worth getting everyone else riled up about. For the past month, I've been trying to solve a RADIUS timeout problem with the HiPer ARC. The majority of this time has been spent with USR techs dialed into my pool verifying that I know what I am talking about, then having them come back and give me workarounds that don't apply (ie: setup a non-RADIUS local user with a timeout). As of today, I still don't have a code update, which is what I presume is the only fix for this problem. My prior ticket on the ARC took a similar amount of time to get a code fix back. This turned out to be a missing conditional according to the engineer who resolved it for me. With Netscape giving out their code, the question I have is, why don't hardware companies do the same? I've already spent the money on the hardware and the software is for all intent purposes, free. I can't take their software and run it on a Portmaster, nor can I use it effectively on a Powermac. However, what I could do with the code is _fix problems_ in a matter of hours, rather than 3Com paying four people to verify the problem, then one person to adjust the code, over a period of a month. The cost effectiveness is readily apparent. The code would also benefit from the Stahlman GNU model, where the more eyes you have looking at the code, the better it becomes. 3Com would hold the base tree, and add adjustments to the source in a submission process. We're not talking about anything that is extremely proprietary. The base ideas behind terminal server hardware are open protocols. I'm certain that if the code was out in the open, 3Com would have all the features that people have been screaming for in a matter of months. If I was shopping around for new hardware, open code would be a BIG selling point.
Subject: (usr-tc) password on netservermanager
From: Heim, Tom J. <tom.heim@teldta.com>
Date: 1998-02-06 14:54:03
Hello, I thought I would try to use the netserver manager program but it asks for a password. I know what the config statement is on the ARC to get HARM working, but is there a similar config on the netserver?? Also, has anyone had trouble upgrading netservers like losing config? thanks, tom.heim@teldta.com
Subject: Re: (usr-tc) HiPer ReBoot
From: Brian <signal@shreve.net>
Date: 1998-02-06 15:03:10
On Fri, 6 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > >> Yes, I see it on 4.0.89. The last technician I talked to about this stated > >> that I should reflash using ZModem. That hasn't help. Looks like I've got > >> to put a logging console on the ARC again. > > > >pete, I opened up a ticket on this #38888, get 4.0.80, its the fix for > >4.0.89 rebooting. I sent 3com my "crashdump" and from that they said I > >needed 4.0.80 and that should fix it. > > 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they number > their engineering releases. me too, I didn't argue it, they sent me 4.0.80 as the fix for 4.0.89 Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Win95 Problem + Code
From: lee@gwinnett.com
Date: 1998-02-06 17:01:56
I just got through helping someone through a problem and was wondering if anyone has seen the same. For the 2nd time in 24 hours, I had a tech support call from a customer who could dial in and authenticate and log in, but could not access anything on our net. The first customer also had NT on his machine, so he tried NT and it worked OK. After checking his settings on Win95, which were OK, I had him go into the TCP/IP Properties inside Network and checked settings. None had to be changed, but I had him click OK instead of cancel on his TCP/IP settings and network properties. The machine made some changes to his settings and asked to reboot (none were changed inside network). After the reboot, I had him dial in and nothing happened. I did suggest to him to remove DUN, TCP/IP, and the winsock.dll, reboot and reinstall DUN. So far, the last I heard from him is it is still not working. The second customer's computer reset the network properties (without changing any of the settings) and requested a reboot. Upon reboot, I had the customer dial in and try to access the net. At first he could not, but after staying connected for two or three minutes, all of a sudden he could access everything. His problem was to my relief, solved. I am wondering if the Netserver card has problems with Win95 DUN. This makes about 30 times I have received requests with help for disconnects, network troubles, etc. from Win95 Customers. I have only seen one or two similar cases with NT (which were quickly solved), and never with Linux. On another note, anyone had any luck getting the Netsever Quake fix 3.6.69? I tried requesting it three days ago, but received no response. Thanks in advance, Lee
Subject: (usr-tc) HiPer DSP oddities...
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-06 17:47:25
Well, finally got around to putting a few PRI spans on HDM cards and playing w/ 'em. An oddity/major concern: Occasionally, for no unknown reason, all the B channels go out-of-service. The D channel looks good, but TCM shows each DS0 Service State as being "remoteOutOfService". Telco says all the channels look good and are idle, yet I cannot receive any calls. The HDM also reports "bad channel id" when an incoming call is attempted, as though the telco is signaling a channel number that the HDM thinks is out of service. Anyone seen anything like this before? Thanks ... Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Open code
From: MegaZone <megazone@livingston.com>
Date: 1998-02-06 18:05:37
Once upon a time Allen Marsalis shaped the electrons to say... >USR (nor any other vendor) would enver go for it. They have a significant >percentage of their total development tied up in software. Who wants guys >like MZ crawling through the "good parts" of the code and fixing their Exactly. Though it wouldn't likely be me, I readily admit I'm not into hardcore code bashing. Scripting and patching, ok. It is really a complete package. This isn't code that runs on open HW. It is custom designed HW and SW that is a SYSTEM! Raw ComOS from a PM-3 won't run on a TC because the HW is radically different, and vice-versa. ComOS running on a MAX would not perform as well as a PM-3, or be able to do some of the tricks, even if optimized for those processors. Why? Becayse the HW architecture is different. We don't design HW and then go write code to run on it. The HW and SW concepts are symbiotic, you design HW for specific tasks and to be able to run the software systems you plan as efficiently as possible. It is a very competitive market. While you can't stop someone from opening a chassis and seeing exactly what ships you use and how they are wired, you can make it damn hard for them to reproduce the *system* by not giving them the source code. If they want to spend all of their time with logic probes trying to sort it out, let them. And they'll forever be WAY behind the times. MS gives away IE because they can *afford to*! They make gobs of money off the rest of their product line. Notice they don't hand out the source to Word, or say Win95, to the general public. NS is giving away the source to Navigator simply to counter MS in the market - and because THEY can afford to. They make most of their money from the servers. Note they aren't giving away the source to all of their servers. Access servers and routers are what we do, and in a fiercely competitive market it would be suicide to open your source. And frankly the users who give a damn are *negligible* so there is no return on doing so. The majority of customers would never look at the source if it were there, and most who would look wouldn't be able to work on it intelligently anyway. Sure, there are a few who might make real improvements or have the ability to make fixes. But that group doesn't begin to register on the market scale involved. And certainly it does not justify the immense competitive risk to having the other vendors in the market getting access to your trade secrets. Sure, it is based on open specs. But that is NOTHING. HOW YOU DO IT is important. If it were nothing, why were people amazed when we did BGP so that you can have multiple feeds in less than 16MB of RAM? Why don't all products have the ability to do DOSBS and modems on the same CT1? Sure, the principles are easy to discuss. Implementing them is another story. -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: Re: (usr-tc) Quake disconnects
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-06 18:39:51
We are getting "quake lag" through Analog (and X2) but not through ISDN. We are running PPP in modem, so I'm not sure if a code update to the NetServer would do the trick or not. Our ISDN is done through the netserver and not in modem. We currently have 3.6.28 on the netserver and 5.5.7/5.6.7 on the quad modems. Any ideas for this setup? Maybe try PPP in NetServer? Or just try to get the NetServer code and see if it works anyway? Thanks for any ideas, Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- >At 03:19 PM 2/3/98 -0600, you wrote: >>At 01:51 PM 2/3/98 -0700, you wrote: >>>Brian Elfert said once upon a time: >>>> >>>>I have a couple of customers complaining that they get disconnected when >>>>playing Quake. >>>> >>>>I'm running 3.6.28 on all of my Netservers. Any ideas? Nobody's >>>>complained about any lag problems to date. >>> >>>Are they using X2? If so, reduce the top speed with &N. >>> >>You may also want to get the 3.6.69 Quake lag fix. >> >This is a fix available via support for Netserver & Quad Modem >combinations.. There are many improvements in the code path for >UDP traffic.. These improvements have also been pushed into the >3.7.x code that is in BETA now.. > >We have mixed reports about improvement.. Some claim their lag went away, >some see improvement but not gone, and others claim that the improvement >is negligible.. If you want to try it, and are experiencing quake lag you >can request it through support or you can wait for the general release of >the 3.7.x code.
Subject: (usr-tc) ARC/HDM problem
From: Brian <signal@shreve.net>
Date: 1998-02-06 19:27:21
I added some hdm's today, and the # of "ports" on them does not look right. Both are configured for PRI, and should show 23 ports: HiPer>> list chassis Slot Owner Description Ports Type . . . 10 YES 24 Channel High Density Modem 23 DYNAMIC 11 YES 24 Channel High Density Modem 23 DYNAMIC 12 YES 24 Channel High Density Modem 23 DYNAMIC 13 YES 24 Channel High Density Modem 23 DYNAMIC 14 YES 24 Channel High Density Modem 24 DYNAMIC 15 YES 24 Channel High Density Modem 0 DYNAMIC as you can see, slot 14 shows 25 and slot 15 shows 0! At least slot 14 will take calls, slot 15 has 0 ports so will just return busies. I *cant* statically set the # of ports, because I have an nmc card in this hub, and its suppose to handle this. All you can really set statically on an arc is card_type and ownership. Anyways, here is what a partial list interfaces shows: . . . slot:14/mod:22 Up Up slot:14/mod:23 Up Up slot:14/mod:24 Down Down internal Up Up loopback Up Up as you can see slot:14 goes to 24......even though it *is* marked "down" I don't like it, because my other HDM's only show 23. Also you can see there is a total absense of slot:15! I have tried to re-insert the card to no avail. If this were a netserver, I could fix it. its almost like I need to do a "set density" or a "set modem active" but I don't know how to do that on a ARC! Any help would be greatly appreciated. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Open code
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-06 19:50:09
At 02:26 PM 2/6/98 -0700, Pete Ashdown ponders: [snip] The code would also benefit from the >Stahlman GNU model, where the more eyes you have looking at the code, the >better it becomes. 3Com would hold the base tree, and add adjustments to >the source in a submission process. > >We're not talking about anything that is extremely proprietary. The base >ideas behind terminal server hardware are open protocols. I'm certain that >if the code was out in the open, 3Com would have all the features that >people have been screaming for in a matter of months. If I was shopping >around for new hardware, open code would be a BIG selling point. > USR (nor any other vendor) would enver go for it. They have a significant percentage of their total development tied up in software. Who wants guys like MZ crawling through the "good parts" of the code and fixing their portmasters. Even MZ would have to admit that there is *something* that TC does better. No code is perfect (unless I write it of course ;) BTW, I love the idea Pete!!! I like the way you think.. allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) ARC/HDM problem
From: Brian <signal@shreve.net>
Date: 1998-02-06 20:16:58
On Fri, 6 Feb 1998, Tatai SV Krishnan wrote: > make sure that you have the chassis awarness enabled - if you have only > one HiPer ARC. > > ena nmc cha > > If you have more than one HiPer ARC then you have to set it static - at > this point - will be fixed soon. > > krish chassis awareness was turned on. I fixed it by doing: set chassis slot 14 carD_TYPE hdm_24 owNER yes typE static ports 23 set chassis slot 15 carD_TYPE hdm_24 owNER yes typE static ports 23 now, its fixed. I was trying to set number of ports while on the phone with tech support (trying to set the num of ports was my idea), and it was complaining that the card was "dynamic", so I asked tech support "but surly there is a way to change it from dynamic to static", and he said "no". So I guess there is no harm in leaving it static, I would like to change it dynamic again if i could and have it retain its # of ports etc. Brian > > > On Fri, 6 Feb 1998, Brian wrote: > > > I added some hdm's today, and the # of "ports" on them does not look > > right. Both are configured for PRI, and should show 23 ports: > > > > HiPer>> list chassis > > Slot Owner Description Ports Type > > . > > . > > . > > 10 YES 24 Channel High Density Modem 23 DYNAMIC > > 11 YES 24 Channel High Density Modem 23 DYNAMIC > > 12 YES 24 Channel High Density Modem 23 DYNAMIC > > 13 YES 24 Channel High Density Modem 23 DYNAMIC > > 14 YES 24 Channel High Density Modem 24 DYNAMIC > > 15 YES 24 Channel High Density Modem 0 DYNAMIC > > > > as you can see, slot 14 shows 25 and slot 15 shows 0! At least slot 14 > > will take calls, slot 15 has 0 ports so will just return busies. > > > > I *cant* statically set the # of ports, because I have an nmc card in this > > hub, and its suppose to handle this. All you can really set statically on > > an arc is card_type and ownership. > > > > Anyways, here is what a partial list interfaces shows: > > . > > . > > . > > slot:14/mod:22 Up Up > > slot:14/mod:23 Up Up > > slot:14/mod:24 Down Down > > internal Up Up > > loopback Up Up > > > > > > as you can see slot:14 goes to 24......even though it *is* marked "down" I > > don't like it, because my other HDM's only show 23. Also you can see > > there is a total absense of slot:15! > > > > I have tried to re-insert the card to no avail. > > > > If this were a netserver, I could fix it. its almost like I need to do a > > "set density" or a "set modem active" but I don't know how to do that on a > > ARC! > > > > Any help would be greatly appreciated. > > > > Brian > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > ----------------------------------------- > \ T.S.V. Krishnan \ > \ Network System Engineer \ ( : - : ) > \ 3Com ............ \ > ----------------------------------------------/ > tkrishna@bubba.ae.usr.com > ----------------------------/ http://interproc.ae.usr.com ----/ > -------------------------------------------------------------------------\ > Any Sufficiently advanced bug is indistinguishable for a feature. > - Rick Kulawiec > -------------------------------------------------------------------------/ > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) ARC/HDM problem
From: Brian <signal@shreve.net>
Date: 1998-02-06 20:16:58
On Fri, 6 Feb 1998, Tatai SV Krishnan wrote: > make sure that you have the chassis awarness enabled - if you have only > one HiPer ARC. > > ena nmc cha > > If you have more than one HiPer ARC then you have to set it static - at > this point - will be fixed soon. > > krish chassis awareness was turned on. I fixed it by doing: set chassis slot 14 carD_TYPE hdm_24 owNER yes typE static ports 23 set chassis slot 15 carD_TYPE hdm_24 owNER yes typE static ports 23 now, its fixed. I was trying to set number of ports while on the phone with tech support (trying to set the num of ports was my idea), and it was complaining that the card was "dynamic", so I asked tech support "but surly there is a way to change it from dynamic to static", and he said "no". So I guess there is no harm in leaving it static, I would like to change it dynamic again if i could and have it retain its # of ports etc. Brian > > > On Fri, 6 Feb 1998, Brian wrote: > > > I added some hdm's today, and the # of "ports" on them does not look > > right. Both are configured for PRI, and should show 23 ports: > > > > HiPer>> list chassis > > Slot Owner Description Ports Type > > . > > . > > . > > 10 YES 24 Channel High Density Modem 23 DYNAMIC > > 11 YES 24 Channel High Density Modem 23 DYNAMIC > > 12 YES 24 Channel High Density Modem 23 DYNAMIC > > 13 YES 24 Channel High Density Modem 23 DYNAMIC > > 14 YES 24 Channel High Density Modem 24 DYNAMIC > > 15 YES 24 Channel High Density Modem 0 DYNAMIC > > > > as you can see, slot 14 shows 25 and slot 15 shows 0! At least slot 14 > > will take calls, slot 15 has 0 ports so will just return busies. > > > > I *cant* statically set the # of ports, because I have an nmc card in this > > hub, and its suppose to handle this. All you can really set statically on > > an arc is card_type and ownership. > > > > Anyways, here is what a partial list interfaces shows: > > . > > . > > . > > slot:14/mod:22 Up Up > > slot:14/mod:23 Up Up > > slot:14/mod:24 Down Down > > internal Up Up > > loopback Up Up > > > > > > as you can see slot:14 goes to 24......even though it *is* marked "down" I > > don't like it, because my other HDM's only show 23. Also you can see > > there is a total absense of slot:15! > > > > I have tried to re-insert the card to no avail. > > > > If this were a netserver, I could fix it. its almost like I need to do a > > "set density" or a "set modem active" but I don't know how to do that on a > > ARC! > > > > Any help would be greatly appreciated. > > > > Brian > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > ----------------------------------------- > \ T.S.V. Krishnan \ > \ Network System Engineer \ ( : - : ) > \ 3Com ............ \ > ----------------------------------------------/ > tkrishna@bubba.ae.usr.com > ----------------------------/ http://interproc.ae.usr.com ----/ > -------------------------------------------------------------------------\ > Any Sufficiently advanced bug is indistinguishable for a feature. > - Rick Kulawiec > -------------------------------------------------------------------------/ > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) New mailing list: nas-talk
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-06 23:00:25
Often the discussions on this list are only tangentially related to US Robotics/3Com Total-Control equipment; sometimes it has to do with vendors, sometimes it has to do with other companies' products, sometimes it's tech support, and sometimes it's other companies' tech support. Personally, I don't like having to sift through all of that to read about the juicy new version of HiPer ARC firmware. There may be others who share that sentiment in the USR community, or similarly-inclined people among the users of equipment by Livingston, Shiva, and others. So I've made the nas-talk mailing list. Please feel free to subscribe, and use it for talking about such things as aren't quite suitable for any one brand of equipment's mailing list. Here's the info file: The nas-talk mailing list is devoted to discussions, comparisons, and critiques of Network Access Servers, and most commonly equipment used to provide dial-in access to networks. nas-talk tends to be unmoderated. nas-talk is for - discussions of dialin equipment, in general - comparison of NAS equipment - relating experiences and testimonials operating NASsen - discussions of manufacturers of NASsen - comparisons of vendors of NASsen nas-talk is not for - information about configuring NAS equipment - advertising - emotional exchanges or pointless arguments To subscribe to nas-talk, email the word "subscribe" to nas-talk-request@datasys.net. The owner of nas-talk can be reached at owner-nas-talk@datasys.net.
Subject: Re: (usr-tc) ISDN/HDM?
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-06 23:17:54
At 11:58 PM 2/6/98 -0500, System Administrator wrote: >Ooof. It's late. I'm tired. Been working for 26 hours now. Pardon any >idiocy or lunatic raving which I might spew forth. > >Can't get ISDN to work over HiPer DSP cards, at all. Analog works fine. >ISDN calls signal good, but the dsp card doesn't seem to know what to do >with it. I've checked and re-checked the configuration 12 million times, >both the hdm card and netserver. > This exact situation occurred to us last week with a new hunt group from a new CLEC. Analog was fine, even x2.. but no idsn. Turns out it was telco related. We have a phone number for every 2 pri's in the hunt. That way we can dial up any chassis we have. And we have a main number that is the access number given out to customers. We were dialing this main number like you would expect. Then we tried one of the pilot numbers and isdn worked great. This lead gave the telco enough information to track down the problem, obviously call- forward related.. Probably not your problem, but I just thought I'd mention it since we had the same exact symptom... allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: (usr-tc) ISDN/HDM?
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-06 23:58:07
Ooof. It's late. I'm tired. Been working for 26 hours now. Pardon any idiocy or lunatic raving which I might spew forth. Can't get ISDN to work over HiPer DSP cards, at all. Analog works fine. ISDN calls signal good, but the dsp card doesn't seem to know what to do with it. I've checked and re-checked the configuration 12 million times, both the hdm card and netserver. Get this in my syslog: Feb 6 21:04:02 tcr1 MODEM: S53: CALL_REF >0x000d0037< PRI_SLOT >0< TS >0< SPAN >0< B_CH >1< Feb 6 21:04:02 tcr1 acct 0x000d0037 dial: S53 call arrived. Feb 6 21:04:02 tcr1 sent out answer incoming call for S53. Feb 6 21:04:07 tcr1 S53 didn't get online! status=-1, connect_fail=22, link_fail=22 Feb 6 21:04:12 tcr1 MODEM: S53: CALL_REF >0x000d0038< PRI_SLOT >0< TS >0< SPAN >0< B_CH >1< So, I'm thinking.. Timeslot 0?!?! Any ideas? Thanks, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) ISDN/HDM?
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-07 00:02:12
At 12:47 AM 2/7/98 -0500, System Administrator wrote: [snip] >I wish these were on a 5ess switch, because I've been through >every imaginable situation w/ circuits off a 5ess, and never seen anything >like this -- so naturally, I assumed it was probably an hdm/configuration >problem. > It was a 5ess switch that our isdn problem occurred on. I've never seen this problem on our dms100 huntgroup.. And both the main and pilot numbers were from the same clec. That is to say, there was no forwarding from one lec/switchtype to another. Just a simple hunt translation on a single 5ess switch.. Actually, we were pleased that we found this so quickly. Our new provider, a clec, was happy to provide more than one number for our pool. BellSouth (gag) was not able to provide that service. Without the extra numbers, we would have had a much harder time tracking it down. allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Open code
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-07 00:12:57
At 06:05 PM 2/6/98 -0800, MegaZone coded in ascii: >And certainly it does not justify the immense competitive >risk to having the other vendors in the market getting access to your >trade secrets. > Hey Zone! Before you leave lucent/livingston, how bout borrowing some portmaster UDP interrupt handlers, and associated procedures and posting the source code here! No, that would be like the defection of the "Red October".. Too strong of a political statement. How about just sending it to a couple of nice usr engineers real secret like... Thanks bud... allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: (usr-tc) exitBeforeX2Connect
From: Matthew Opoka <phantom@magnolia.net>
Date: 1998-02-07 00:45:44
What does exitBeforeX2Connect mean?
Subject: Re: (usr-tc) ISDN/HDM?
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-07 00:47:39
On Fri, 6 Feb 1998, Allen Marsalis wrote: > At 11:58 PM 2/6/98 -0500, System Administrator wrote: > >Ooof. It's late. I'm tired. Been working for 26 hours now. Pardon any > >idiocy or lunatic raving which I might spew forth. > > > >Can't get ISDN to work over HiPer DSP cards, at all. Analog works fine. > >ISDN calls signal good, but the dsp card doesn't seem to know what to do > >with it. I've checked and re-checked the configuration 12 million times, > >both the hdm card and netserver. > > > > This exact situation occurred to us last week with a new hunt group > from a new CLEC. Analog was fine, even x2.. but no idsn. Turns out > it was telco related. We have a phone number for every 2 pri's in > the hunt. That way we can dial up any chassis we have. And we have > a main number that is the access number given out to customers. We > were dialing this main number like you would expect. Then we tried > one of the pilot numbers and isdn worked great. This lead gave the > telco enough information to track down the problem, obviously call- > forward related.. > > Probably not your problem, but I just thought I'd mention it since > we had the same exact symptom... Actually, that's definitely something I'm going to look into. We just had three new pri circuits installed today.. Two of 'em are up, telco has some sort of cross-connect problem w/ the "middle" circuit at their tandem. Normally, I would have test numbers for each circuit, however our "coordinator" at the telco screwed the pooch and didn't write up the order properly (none of this amazes anyone on this list of course, it is after all ... a telephone company), so all I have for testing right now is a single "office route" DN (what dms100 geeks call phone numbers assigned to hunt groups) which is supposed to start at the first circuit and roll downward. I should have individual test numbers for each circuit on Monday, so I'll probably just put it on hold until then and see what happens. I wish these were on a 5ess switch, because I've been through every imaginable situation w/ circuits off a 5ess, and never seen anything like this -- so naturally, I assumed it was probably an hdm/configuration problem. Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: (usr-tc) blocking non x2 subscribers...
From: Matthew Opoka <phantom@magnolia.net>
Date: 1998-02-07 00:49:21
We have a TC with 48 digital/analog modems the first half we are using regular phone lines on the other half we are using a PRI. To authenticate the user we are using Livingstons Radius 2.01. How can I set it so only X2 subscribers can use the digital lines and not everyone? Thanks Matthew
Subject: Re: (usr-tc) HiPer DSP oddities...
From: matthew <matthew@the-spa.com>
Date: 1998-02-07 01:02:31
no. never seen ANY problems with our double up cards yet. i had really put off putting in the first set because i was expecting a bunch of problems just because they are so new but no problems at all. is this happening with more than one set or just one chassis or even just one card? have you tried reflashing all the code? matthew At 05:47 PM 2/6/98 -0500, you wrote: >Well, finally got around to putting a few PRI spans on HDM cards and >playing w/ 'em. An oddity/major concern: > >Occasionally, for no unknown reason, all the B channels go out-of-service. >The D channel looks good, but TCM shows each DS0 Service State as being >"remoteOutOfService". Telco says all the channels look good and are idle, >yet I cannot receive any calls. The HDM also reports "bad channel id" >when an incoming call is attempted, as though the telco is signaling a >channel number that the HDM thinks is out of service. > >Anyone seen anything like this before? > >Thanks ... > >Jesse Sipprell >Senior Systems Engineer >Evolution Communications, Inc. > >* Finger sysadmin@evcom.net for my PGP Public Key * > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) New mailing list: nas-talk
From: MegaZone <megazone@livingston.com>
Date: 1998-02-07 03:25:42
Once upon a time Mark R. Lindsey shaped the electrons to say... >So I've made the nas-talk mailing list. Please feel free to subscribe, >and use it for talking about such things as aren't quite suitable for >any one brand of equipment's mailing list. So, another to add to the list of inet-access, IAP, isp-tech (TWO lists by that name), etc... What was wrong with the existing lists? BTW, it won't work. People have been adding more and more 'general' lists and it doesn't reduce the traffic on others. And now I see a lot of cross-posting... -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: Re: (usr-tc) hiperarc question?
From: MyungSik Kim <daesung@soback.kornet.nm.kr>
Date: 1998-02-07 10:34:18
Pete, This is what I want. If user use terminal mode, they just connected unix host without asking userid&password (also unix host doesn't ask userid & password because its hosts.equiv file has NAS ip address) or ppp then they just be authenticated. Does anyone implement service like this? On Fri, 6 Feb 1998, Pete Ashdown wrote: > MyungSik Kim said once upon a time: > > >Can you help me how to set rlogin -l up ? Regards > > Here are sample settings for our "guest" user on the HiPer ARC: > > add user guest password "" login_service rlogin type login > set login user guest login_host_ip_address 198.60.22.2 > set login user guest terminal_type dialup > enable user guest >
Subject: (usr-tc) exitBeforeX2Connect
From: Rick Payne <rickp@corp.netcom.net.uk>
Date: 1998-02-07 10:54:11
>>>>> "Matthew" == Matthew Opoka <phantom@magnolia.net> writes: Matthew> What does exitBeforeX2Connect mean? It my experience the chassis reports this when a call drops out of X2, or fails to achieve an X2 connect in the first place (even though both ends are X2 capable). Rick -- Rick Payne, Senior Network Admin | Taxation is about extracting the NETCOM Internet Ltd. | maximum amount of milk with the rickp@corp.netcom.net.uk | minimum of moo.
Subject: Re: (usr-tc) HiPer DSP oddities...
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-07 11:03:00
On Sat, 7 Feb 1998, matthew wrote: > no. never seen ANY problems with our double up cards yet. i had really put > off putting in the first set because i was expecting a bunch of problems > just because they are so new but no problems at all. > > is this happening with more than one set or just one chassis or even just > one card? It's happened a couple times to multiple cards in multiple chassis. > have you tried reflashing all the code? Not yet. I'll do so if I see it happen again. > matthew > > At 05:47 PM 2/6/98 -0500, you wrote: > >Well, finally got around to putting a few PRI spans on HDM cards and > >playing w/ 'em. An oddity/major concern: > > > >Occasionally, for no unknown reason, all the B channels go out-of-service. > >The D channel looks good, but TCM shows each DS0 Service State as being > >"remoteOutOfService". Telco says all the channels look good and are idle, > >yet I cannot receive any calls. The HDM also reports "bad channel id" > >when an incoming call is attempted, as though the telco is signaling a > >channel number that the HDM thinks is out of service. > > > >Anyone seen anything like this before? > > > >Thanks ... Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) ISDN/HDM?
From: Brian <signal@shreve.net>
Date: 1998-02-07 11:20:28
On Fri, 6 Feb 1998, System Administrator wrote: > Ooof. It's late. I'm tired. Been working for 26 hours now. Pardon any > idiocy or lunatic raving which I might spew forth. > > Can't get ISDN to work over HiPer DSP cards, at all. Analog works fine. > ISDN calls signal good, but the dsp card doesn't seem to know what to do > with it. I've checked and re-checked the configuration 12 million times, > both the hdm card and netserver. Ok, post your ISDN Call Control Options...... > > Get this in my syslog: > > Feb 6 21:04:02 tcr1 MODEM: S53: CALL_REF >0x000d0037< PRI_SLOT >0< TS >0< > SPAN >0< B_CH >1< > Feb 6 21:04:02 tcr1 acct 0x000d0037 dial: S53 call arrived. > Feb 6 21:04:02 tcr1 sent out answer incoming call for S53. > Feb 6 21:04:07 tcr1 S53 didn't get online! status=-1, connect_fail=22, > link_fail=22 > Feb 6 21:04:12 tcr1 MODEM: S53: CALL_REF >0x000d0038< PRI_SLOT >0< TS >0< > SPAN >0< B_CH >1< > > So, I'm thinking.. Timeslot 0?!?! > > Any ideas? > > Thanks, > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) ISDN/HDM?
From: Brian <signal@shreve.net>
Date: 1998-02-07 11:20:28
On Fri, 6 Feb 1998, System Administrator wrote: > Ooof. It's late. I'm tired. Been working for 26 hours now. Pardon any > idiocy or lunatic raving which I might spew forth. > > Can't get ISDN to work over HiPer DSP cards, at all. Analog works fine. > ISDN calls signal good, but the dsp card doesn't seem to know what to do > with it. I've checked and re-checked the configuration 12 million times, > both the hdm card and netserver. Ok, post your ISDN Call Control Options...... > > Get this in my syslog: > > Feb 6 21:04:02 tcr1 MODEM: S53: CALL_REF >0x000d0037< PRI_SLOT >0< TS >0< > SPAN >0< B_CH >1< > Feb 6 21:04:02 tcr1 acct 0x000d0037 dial: S53 call arrived. > Feb 6 21:04:02 tcr1 sent out answer incoming call for S53. > Feb 6 21:04:07 tcr1 S53 didn't get online! status=-1, connect_fail=22, > link_fail=22 > Feb 6 21:04:12 tcr1 MODEM: S53: CALL_REF >0x000d0038< PRI_SLOT >0< TS >0< > SPAN >0< B_CH >1< > > So, I'm thinking.. Timeslot 0?!?! > > Any ideas? > > Thanks, > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: Ian Roy <iroy@smokies.amnix.com>
Date: 1998-02-07 11:49:15
> We have a TC with 48 digital/analog modems the first half we are using > regular phone lines on the other half we are using a PRI. > To authenticate the user we are using Livingstons Radius 2.01. > How can I set it so only X2 subscribers can use the digital lines and not > everyone? > > Thanks > > Matthew > > Matthew, We were looking for a solution to the same problem. We use Merit Radius here. One of the options we considered was to hack Merit Radius to check for this, but gave up on it. The solution we are using - We hacked together a program that does the following: a. Runs every minute from as a cron job b. Reads a configuration file of the services a subscriber has subscribed to (not mentioned in the file means a 33.6K service) c. Telnets to the USRTCs, and the other dialup routers we have on our network d. Gets a list of who is logged in on which router and which port e. Checks to see whether this user is allowed to be using the particular service f. If not, it kicks them off the port they are on g. Logs the fact into a log file It seems to be working pretty well for us. After we put it into production use, we received numerous calls from subscribers saying they are getting disconnected. For each such complaint, we check against the logs and tell them they are trying to use a service they are not signed up for. "Would you like to signup for that service ?" It seems to work pretty good for us. Thanks Ian Roy
Subject: Re: (usr-tc) Open code
From: Carl Ansley <carl@caverock.co.nz>
Date: 1998-02-07 13:24:03
> We're not talking about anything that is extremely proprietary. The base > ideas behind terminal server hardware are open protocols. I'm certain that > if the code was out in the open, 3Com would have all the features that > people have been screaming for in a matter of months. If I was shopping > around for new hardware, open code would be a BIG selling point. Agreed. Up until the point we bought the USR/3com hardware, _all_ the software we used on our servers we had the source code to. The reason we use Linux, Apache, squid and so forth is because we have the source code and can modify functionality, fix problems and add features as we desire. Even with commercial stuff, our policy is to only buy server software that comes with source code. The TC is the only exception to this rule at the moment. carl -- Carl Ansley (carl@caverock.co.nz) Phone: +64 3 3664242 Cave Rock Software / Cave Rock Internet Fax: +64 3 3665478 Unit 1b, 492 Moorhouse Ave, PO Box 22488, Christchurch, New Zealand
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Randy Doran <rtdoran@gate.net>
Date: 1998-02-07 15:35:45
On Mon, 2 Feb 1998, Pete Ashdown wrote: > Charles Sprickman said once upon a time: > > >I have something of a follow-on question... We're currently using an > >older Merit Radius server (2.23.24?) which works, but I feel like I'm > >missing out on some goodies. I'm looking to either upgrade to the > >Livingston code or the newest Merit release, but I'm a bit confused about > >the vendor-specific attributes and how one goes about "merging" the USR > >dictionary in. > > Someone posted a patch a month or so ago. I haven't implemented it yet, so > I can't comment on it. > > >Also, does any (free) Radius server support port-limits? We have source > >to the USR radiusd (don't ask) but it supposedly requires some external > >pay-to-play stuff to actually compile... > > The Merit does, use "Port-Limit". This only applies to single chassis > though. Does this mean that a user set to Port-Limit=1 can still get 2 "B" channels through MPIP? Randy Doran
Subject: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-07 21:29:50
I have a subnet, declared as follows: signal Authentication-Type = Unix-PW Service-Type = Framed, Framed-Protocol = PPP, Framed-IP-Address = 208.214.45.1, Framed-Route = "208.214.45.0/28 208.214.45.1 1", Framed-MTU = 1500, Framed-Compression = Van-Jacobson-TCP-IP yet, whenever I start gated on a machine (running rip), I get this in the routing tables: 208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 enh_routing is on broadcast/listen is on net0 as well its almost like the netserver is sending out a v1 route, or maybe gated is interpreting a v2 route as v1, but I dont think so since all other parts of 208.214.45 (all /28's) work fine. Anyone know why this would happen? only with that .0 subnet does this happen. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Brian <signal@shreve.net>
Date: 1998-02-07 21:35:57
On Sat, 7 Feb 1998, Randy Doran wrote: > > On Mon, 2 Feb 1998, Pete Ashdown wrote: > > > Charles Sprickman said once upon a time: > > > > >I have something of a follow-on question... We're currently using an > > >older Merit Radius server (2.23.24?) which works, but I feel like I'm > > >missing out on some goodies. I'm looking to either upgrade to the > > >Livingston code or the newest Merit release, but I'm a bit confused about > > >the vendor-specific attributes and how one goes about "merging" the USR > > >dictionary in. > > > > Someone posted a patch a month or so ago. I haven't implemented it yet, so > > I can't comment on it. > > > > >Also, does any (free) Radius server support port-limits? We have source > > >to the USR radiusd (don't ask) but it supposedly requires some external > > >pay-to-play stuff to actually compile... > > > > The Merit does, use "Port-Limit". This only applies to single chassis > > though. > > Does this mean that a user set to Port-Limit=1 can still get 2 "B" > channels through MPIP? yes, so run something like PMMON(tsmon) www.tsmon.com Brian > > Randy Doran > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) HiPer ReBoot
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-07 22:57:45
> 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they number > their engineering releases. Normal releases work upwards. Engineering releases work backwards. Why? Heaven only knows... Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-07 23:13:27
> > > > enh_routing is on > > broadcast/listen is on net0 as well > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > interpreting a v2 route as v1, but I dont think so since all other parts > > of 208.214.45 (all /28's) work fine. > > > > > > set ripv2 on > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver from sending out any v1 info? perhaps my cisco is getting the v2 data and sending out v1 as well? Brian > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-07 23:13:27
> > > > enh_routing is on > > broadcast/listen is on net0 as well > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > interpreting a v2 route as v1, but I dont think so since all other parts > > of 208.214.45 (all /28's) work fine. > > > > > > set ripv2 on > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver from sending out any v1 info? perhaps my cisco is getting the v2 data and sending out v1 as well? Brian > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-07 23:58:15
On Sat, 7 Feb 1998, Brian wrote: > > I have a subnet, declared as follows: > > signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > > yet, whenever I start gated on a machine (running rip), I get this in the > routing tables: > > 208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 > > > enh_routing is on > broadcast/listen is on net0 as well > > its almost like the netserver is sending out a v1 route, or maybe gated is > interpreting a v2 route as v1, but I dont think so since all other parts > of 208.214.45 (all /28's) work fine. > > set ripv2 on
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-07 23:58:15
On Sat, 7 Feb 1998, Brian wrote: > > I have a subnet, declared as follows: > > signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > > yet, whenever I start gated on a machine (running rip), I get this in the > routing tables: > > 208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 > > > enh_routing is on > broadcast/listen is on net0 as well > > its almost like the netserver is sending out a v1 route, or maybe gated is > interpreting a v2 route as v1, but I dont think so since all other parts > of 208.214.45 (all /28's) work fine. > > set ripv2 on
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-08 02:01:12
On Sat, 7 Feb 1998, Brian wrote: > > > > > > enh_routing is on > > > broadcast/listen is on net0 as well > > > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > > interpreting a v2 route as v1, but I dont think so since all other parts > > > of 208.214.45 (all /28's) work fine. > > > > > > > > > > set ripv2 on > > > > > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver > from sending out any v1 info? perhaps my cisco is getting the v2 data and > sending out v1 as well? On cisco you have to enable ripv2 as well and that is only available on IOS 11.1 and higher.. Prior to that ciscos did not support v2...
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-08 02:01:12
On Sat, 7 Feb 1998, Brian wrote: > > > > > > enh_routing is on > > > broadcast/listen is on net0 as well > > > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > > interpreting a v2 route as v1, but I dont think so since all other parts > > > of 208.214.45 (all /28's) work fine. > > > > > > > > > > set ripv2 on > > > > > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver > from sending out any v1 info? perhaps my cisco is getting the v2 data and > sending out v1 as well? On cisco you have to enable ripv2 as well and that is only available on IOS 11.1 and higher.. Prior to that ciscos did not support v2...
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-08 12:24:17
On Sun, 8 Feb 1998, Garry Shtern wrote: > > > On Sat, 7 Feb 1998, Brian wrote: > > > > > > > > > enh_routing is on > > > > broadcast/listen is on net0 as well > > > > > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > > > interpreting a v2 route as v1, but I dont think so since all other parts > > > > of 208.214.45 (all /28's) work fine. > > > > > > > > > > > > > > set ripv2 on > > > > > > > > > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver > > from sending out any v1 info? perhaps my cisco is getting the v2 data and > > sending out v1 as well? > > On cisco you have to enable ripv2 as well and that is only available on > IOS 11.1 and higher.. Prior to that ciscos did not support v2... > Understood: router rip version 2 network 208.206.76.0 timers basic 30 30 2 60 300 but, does my above config *disable* v1, or merely turn on v2? I want my network to have NO v1, just v2. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-08 12:24:17
On Sun, 8 Feb 1998, Garry Shtern wrote: > > > On Sat, 7 Feb 1998, Brian wrote: > > > > > > > > > enh_routing is on > > > > broadcast/listen is on net0 as well > > > > > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > > > interpreting a v2 route as v1, but I dont think so since all other parts > > > > of 208.214.45 (all /28's) work fine. > > > > > > > > > > > > > > set ripv2 on > > > > > > > > > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver > > from sending out any v1 info? perhaps my cisco is getting the v2 data and > > sending out v1 as well? > > On cisco you have to enable ripv2 as well and that is only available on > IOS 11.1 and higher.. Prior to that ciscos did not support v2... > Understood: router rip version 2 network 208.206.76.0 timers basic 30 30 2 60 300 but, does my above config *disable* v1, or merely turn on v2? I want my network to have NO v1, just v2. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver route problem
From: Wayne Barber <barberw@tidewater.net>
Date: 1998-02-08 15:00:52
> Date: Sun, 8 Feb 1998 12:24:17 -0600 (CST) > From: Brian <signal@shreve.net> > To: usr-tc@lists.xmission.com > Cc: USRobotics TC Mailing List <usr-tc@xmission.com> > Subject: Re: (usr-tc) Netserver route problem Brian, I wish I could help with your problem or answer your questions, but I can't. However, would it be possible to get you to stop ccing the list? It causes everyone to receive your messages twice (and also doubles the replies from some people). Thanks, Wayne Barber - barberw@tidewater.net Internet System Administrator Coastal Telco Services
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-08 16:01:39
On Sun, 8 Feb 1998, Wayne Barber wrote: > > Date: Sun, 8 Feb 1998 12:24:17 -0600 (CST) > > From: Brian <signal@shreve.net> > > To: usr-tc@lists.xmission.com > > Cc: USRobotics TC Mailing List <usr-tc@xmission.com> > > Subject: Re: (usr-tc) Netserver route problem > > Brian, > I wish I could help with your problem or answer your questions, but I > can't. However, would it be possible to get you to stop ccing the > list? It causes everyone to receive your messages twice (and also > doubles the replies from some people). > yes, sorry about that, has to do with the way the headers are on the usr-tc list vs the way my email program replies to things by default, I will try and watch that. Brian > Thanks, > > Wayne Barber - barberw@tidewater.net > Internet System Administrator > Coastal Telco Services > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-08 20:27:24
On Mon, 9 Feb 1998, Bob Purdon wrote: > > > router rip > > version 2 > > network 208.206.76.0 > > timers basic 30 30 2 60 300 > > > > but, does my above config *disable* v1, or merely turn on v2? I want my > > network to have NO v1, just v2. > > Nope, that (if my reading is correct) will support v1 and v2 on all > interfaces. The send/receive versions are set on a per-interface basis: > > ip rip send version 1|2 > ip rip receive version 1|2 > Right, and I think that v1 is declared by "router rip" just by default. so what you think I need to do something like int e0 no ip rip send version 1 no ip rip receive version 1 ip rip send version 2 ip rip receive version 2 > Cheers, > > Bob. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-09 00:49:45
On Sun, 8 Feb 1998, Brian wrote: > On Sun, 8 Feb 1998, Garry Shtern wrote: > > > > > > > On Sat, 7 Feb 1998, Brian wrote: > > > > > > > > > > > > enh_routing is on > > > > > broadcast/listen is on net0 as well > > > > > > > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > > > > interpreting a v2 route as v1, but I dont think so since all other parts > > > > > of 208.214.45 (all /28's) work fine. > > > > > > > > > > > > > > > > > > set ripv2 on > > > > > > > > > > > > > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver > > > from sending out any v1 info? perhaps my cisco is getting the v2 data and > > > sending out v1 as well? > > > > On cisco you have to enable ripv2 as well and that is only available on > > IOS 11.1 and higher.. Prior to that ciscos did not support v2... > > > > Understood: > > router rip > version 2 > network 208.206.76.0 > timers basic 30 30 2 60 300 > > > but, does my above config *disable* v1, or merely turn on v2? I want my > network to have NO v1, just v2. > that is precisely what it does... It disables v1 completely..
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-09 00:49:45
On Sun, 8 Feb 1998, Brian wrote: > On Sun, 8 Feb 1998, Garry Shtern wrote: > > > > > > > On Sat, 7 Feb 1998, Brian wrote: > > > > > > > > > > > > enh_routing is on > > > > > broadcast/listen is on net0 as well > > > > > > > > > > its almost like the netserver is sending out a v1 route, or maybe gated is > > > > > interpreting a v2 route as v1, but I dont think so since all other parts > > > > > of 208.214.45 (all /28's) work fine. > > > > > > > > > > > > > > > > > > set ripv2 on > > > > > > > > > > > > > > ripv2 is on, hmm, when you turn ripv2 on, does that stop the netserver > > > from sending out any v1 info? perhaps my cisco is getting the v2 data and > > > sending out v1 as well? > > > > On cisco you have to enable ripv2 as well and that is only available on > > IOS 11.1 and higher.. Prior to that ciscos did not support v2... > > > > Understood: > > router rip > version 2 > network 208.206.76.0 > timers basic 30 30 2 60 300 > > > but, does my above config *disable* v1, or merely turn on v2? I want my > network to have NO v1, just v2. > that is precisely what it does... It disables v1 completely..
Subject: Re: (usr-tc) livingston radius non-functional
From: MegaZone <megazone@livingston.com>
Date: 1998-02-09 01:06:10
Once upon a time Rev. Jim shaped the electrons to say... >i just installed a NetServer and am trying to get radius to work - i >downloaded the 2.0.1 from livingston and have it sort of running - i get I hope you have some Lucent/Livingston HW. Check the license agreement that comes with RADIUS, it is authorized only for owners of PortMasters. It is in the file called, of all things, 'LICENSE' when you untar the download. ---excerpt--- (a) such use of the Software must be made in conjunction with Livingston Products, ---excerpt--- >radiusd: Livingston RADIUS 2.0.1 97/5/22 NDBM NOSHADOW PASSCHANGE flat_users > >it would appear that 2.0.1 was compiled without shadow support - i have No, that means it was copiled with -DNOSHADOW, which does not mean no shadow passwords, it means no shadow.h file. Most systems that do shadow passwords do not use shadow.h, but a few do. If you need that file compiled in, then you do need to recomlike. >i have been reasonably diligent on searching the livingston site for the >path/src directory and have come up short - let me guess, i have to by one >of livingstons boxes to get the source to the latest radius Ding! And this sounds like an admission that you don't own a PortMaster, which means you are using RADIUS 2.0.1 illegally. Kindly delete it immediately. Any Livingston/Lucent RADIUS 2.0 or later is only legally usable by PortMaster owners. If you own a PM you can use it with any vendors HW, but you have to have at least one piece of our HW. -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: (usr-tc) livingston radius non-functional
From: Rev. Jim <packrat@aus-etc.com>
Date: 1998-02-09 01:40:09
i just installed a NetServer and am trying to get radius to work - i downloaded the 2.0.1 from livingston and have it sort of running - i get /var/adm/radacct/netserver/detail entries when i create a user in the NetServer GUI wizard and log in but cannot get an authentication from my linux 2.0.30 passwd/shadow radius server from: gandalf:# radiusd -v radiusd: Livingston RADIUS 2.0.1 97/5/22 NDBM NOSHADOW PASSCHANGE flat_users it would appear that 2.0.1 was compiled without shadow support - i have seen mention in this list of changing the flag -DSHADOW and re-compiling - i have been reasonably diligent on searching the livingston site for the path/src directory and have come up short - let me guess, i have to by one of livingstons boxes to get the source to the latest radius i am open to suggestions now on which way to turn tia for any assistance rendered If your new sofa was on a curb in another part of town yesterday... you might be a redneck. packrat
Subject: Re: (usr-tc) livingston radius non-functional
From: Jay M Christner <jaymc@goshen.edu>
Date: 1998-02-09 08:30:28
Hmmmm, I was told (by someone AT LIVINGSTON) that it was okay to use the server (RadiusNT 2.0 was in question at the time) with other equipment, but it was completely unsupported. Seems sorta contradictory. Anyway, it's a moot point for me now, but there seems to be some dissention in the ranks. :) -j At 01:06 AM 2/9/98 -0800, you wrote: >Once upon a time Rev. Jim shaped the electrons to say... >>i just installed a NetServer and am trying to get radius to work - i >>downloaded the 2.0.1 from livingston and have it sort of running - i get > >I hope you have some Lucent/Livingston HW. Check the license agreement >that comes with RADIUS, it is authorized only for owners of PortMasters. >It is in the file called, of all things, 'LICENSE' when you untar the >download. > >---excerpt--- >(a) such use of the Software must be made in conjunction with >Livingston Products, >---excerpt--- > >>radiusd: Livingston RADIUS 2.0.1 97/5/22 NDBM NOSHADOW PASSCHANGE flat_users >> >>it would appear that 2.0.1 was compiled without shadow support - i have > >No, that means it was copiled with -DNOSHADOW, which does not mean no >shadow passwords, it means no shadow.h file. Most systems that do >shadow passwords do not use shadow.h, but a few do. If you need that >file compiled in, then you do need to recomlike. > >>i have been reasonably diligent on searching the livingston site for the >>path/src directory and have come up short - let me guess, i have to by one >>of livingstons boxes to get the source to the latest radius > >Ding! And this sounds like an admission that you don't own a PortMaster, >which means you are using RADIUS 2.0.1 illegally. Kindly delete it >immediately. Any Livingston/Lucent RADIUS 2.0 or later is only legally >usable by PortMaster owners. If you own a PM you can use it with any >vendors HW, but you have to have at least one piece of our HW. > >-MZ >-- >Lucent Remote Access Division - Chair, Department of Interstitial Affairs >Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com >For support requests: support@livingston.com <http://www.livingston.com/> >Snail mail: 4464 Willow Road, Pleasanton, CA 94588 > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: RE: (usr-tc) HiPer ReBoot
From: Gillett, Doug L. <doug.gillett@teldta.com>
Date: 1998-02-09 08:33:26
> >>> 4.0.80 is a fix for 4.0.89?? Now I'm really clueless as to how they >>>number >>> their engineering releases. > >> Normal releases work upwards. Engineering releases work backwards. Why? >> Heaven only knows... > >> Regards, > >> Bob Purdon Can anyone confirm that 4.0.80 does indeed solve the problem? Also, if someone could provide me some information regarding the types of scenarios under which the reboot occurs I would appreciate it. I have a HiPer system just in from 3Com for evaluation and would like to reproduce the problem. Thanks, --dlg -- Doug Gillett >doug.gillett@teldta.com
Subject: Re: (usr-tc) livingston radius non-functional
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-09 08:49:37
On Mon, 9 Feb 1998, MegaZone wrote: Speaking of RADIUS, I plan on completely redeveloping redesigning our internal radius server (currently highly hacked). When I get done, I'm going to GPL it and release it. I won't actually be able to *support* it, however those who find bugs and fix them are welcome to send patches to me, and I'll maintain a source tree. I will, of course, fix/distribute any bugs that directly affect my personal usage. :)) Features: * SQL support (currently: postgresql and msql). * Ability to plug in modular scripts (perl, tcl, etc) in order to extend data storage capability to any backend SQL server or database format. * Capable of reading CHAP required passwords from a non-cleartext format (using internal md5 two-way encryption based on a source configurable secret). Two-way encryption is not completely secure, but it does mean that any evil-doers will need to either (a) brute force the encryption or (b) get a copy of the source secret in order to determine user passwords. * Can read unix style password/shadow files from alternate sources, so that every radius account need not be in system wide authentication databases. * Port-Limit can be configured to monitor multiple NASen and prohibit exceeding limits across multiple boxes. * Full syslog logging, or SQL logging. * Special logs stored (or SQL) for current connections, and last connection list. Remote (authorized) access to these provided by two administrative tcp ports and scripts, or via perl scripts which query SQL tables. Regards, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: (usr-tc) Power Supply
From: Eric J. Merkel <merkel@defnet.com>
Date: 1998-02-09 10:05:14
I am wanting to get a second power supply for our TC hub. Redundant power is always a good thing (TM). :) Anybody know about how much a 70A PSU runs for a TC rack? What's a good source for new/used TC equipment like this? Eric ============================================================================= Eric Merkel | URL: www.metalink.net | Local Access in MetaLink Technologies, Inc | EMail: merkel@defnet.com | Defiance, Fulton, 419-782-3472 Ext. 4 | Sales: 1-888-999-8002 | Henry, & Williams Co. =============================================================================
Subject: (usr-tc) HiPerARC reboots ??????
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-09 10:36:14
I've had some bad surprises over the week-end, one of my HiPerARC's rebooted 3 times for no apparent reasons. The only message I had in the CRITICAL EVENTS table is the following (same thing in syslog) :=20 At 19:42:33, Facility "IP", Level "CRITICAL":: No IP Address Pools created At 19:42:33, Facility "IP", Level "CRITICAL":: ip_fwd_get_opt: no IP address available for dynamic address assignment Even though I have an IP pool, and the card takes calls, authentifies and assigns IP addresses. An other weird thing, the card shows up as yellow in the TCM, with Run/Fail LED blinking green=A0?=A0?=A0?=A0? Does that mean the card is = smoked, even though it seems to work perfectly. I'm beginning to suspect the PRI's here, as I have a Cisco 3640 with dual PRI card which got stuck as well for no apparent reason, and we traced the problem to the PRI card. We swapped it for a new one, didn't help. Could the instability of a PRI be the cause of those problems=A0? = Thanks for any info, Robert von Bismarck Petrel Communications SA=20
Subject: Re: (usr-tc) Netserver route problem
From: Jeff Lynch <jeff@mercury.jorsm.com>
Date: 1998-02-09 11:03:58
On Sun, 8 Feb 1998, Wayne Barber wrote: > > Date: Sun, 8 Feb 1998 12:24:17 -0600 (CST) > > From: Brian <signal@shreve.net> > > To: usr-tc@lists.xmission.com > > Cc: USRobotics TC Mailing List <usr-tc@xmission.com> > > Subject: Re: (usr-tc) Netserver route problem > > Brian, > I wish I could help with your problem or answer your questions, but I > can't. However, would it be possible to get you to stop ccing the > list? It causes everyone to receive your messages twice (and also > doubles the replies from some people). > > Thanks, > > Wayne Barber - barberw@tidewater.net > Internet System Administrator > Coastal Telco Services I do not concur with this request. Brian is working out a problem that could very easily help the rest of us when if we have the same problem. If this was a simple config issue the total solution would have been posted and completed...as I'm certain will eventually happen. In that case, I will be happy to see it on this list. Dup replies are from people not being careful about who is in the cc and to lists before sending it. ========================================================================= Jeffrey A. Lynch, President JORSM Internet email: jeff@jorsm.com Northwest Indiana's Full-Service Provider Voice: (219)322-2180 927 Sheffield Avenue, Dyer, IN 46311 Autoresponse: info@jorsm.com http://www.jorsm.com
Subject: Re: (usr-tc) Power Supply
From: Brian <signal@shreve.net>
Date: 1998-02-09 11:32:21
On Mon, 9 Feb 1998, Eric J. Merkel wrote: > I am wanting to get a second power supply for our TC hub. > Redundant power is always a good thing (TM). :) > > Anybody know about how much a 70A PSU runs for a TC rack? > What's a good source for new/used TC equipment like this? > > Eric Call Source Technology, but I would go for the 130A just in case you ever grow enough to put 14HDM's in a chassis. Brian > > > > > > > ============================================================================= > Eric Merkel | URL: www.metalink.net | Local Access in > MetaLink Technologies, Inc | EMail: merkel@defnet.com | Defiance, Fulton, > 419-782-3472 Ext. 4 | Sales: 1-888-999-8002 | Henry, & Williams Co. > ============================================================================= > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Help ! Radius not accurate
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 12:10:13
Randy Doran said once upon a time: >> The Merit does, use "Port-Limit". This only applies to single chassis >> though. > >Does this mean that a user set to Port-Limit=1 can still get 2 "B" >channels through MPIP? Actually no. However, if you're talking about multiple logins, people can share an account on multiple chassis, even if Port-Limit is set to 1.
Subject: Re: (usr-tc) PSU in alarm
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 12:36:21
Brian said once upon a time: > >One of my PSU's is in alarm.............I suppose its toast, but is there >anything I should try? > >1. I have tried reinserting it (powered off to do this of course) >2. I have checked the one fuse on the PSU > >Any other suggestions? Is this common? Send it back. I've had two go bad. The second was due to a nasty wiring fault, but the first went on its own.
Subject: Re: (usr-tc) Netserver route problem
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-09 12:44:04
> router rip > version 2 > network 208.206.76.0 > timers basic 30 30 2 60 300 > > but, does my above config *disable* v1, or merely turn on v2? I want my > network to have NO v1, just v2. Nope, that (if my reading is correct) will support v1 and v2 on all interfaces. The send/receive versions are set on a per-interface basis: ip rip send version 1|2 ip rip receive version 1|2 Cheers, Bob.
Subject: (usr-tc) PSU in alarm
From: Brian <signal@shreve.net>
Date: 1998-02-09 12:48:43
One of my PSU's is in alarm.............I suppose its toast, but is there anything I should try? 1. I have tried reinserting it (powered off to do this of course) 2. I have checked the one fuse on the PSU Any other suggestions? Is this common? Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Open code
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 12:51:33
Allen Marsalis said once upon a time: >USR (nor any other vendor) would enver go for it. They have a significant >percentage of their total development tied up in software. Who wants guys >like MZ crawling through the "good parts" of the code and fixing their >portmasters. Even MZ would have to admit that there is *something* that >TC does better. No code is perfect (unless I write it of course ;) > >BTW, I love the idea Pete!!! I like the way you think.. The idea of competitors stealing code blows the GNU and the Netscape model. The GNU license clearly states that you are free to use the code in any way, just as long as you redistribute the source with it. If Livingston was to rape USR's code (or vice-versa) for neat-stuff, then they would be forced into redistributing their source as well. Otherwise one could sue the pants off the other. Don't you think Netscape is thinking about Microsoft when they're putting out source to Navigator? The only reason open code will probably never see the light of day is the fact that the suits don't understand the concept. They'd rather be protectionary and hostile towards any outside eyes friendly or not. In the process they'll be dumping zillions on development and support.
Subject: Re: (usr-tc) hiperarc question?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 12:52:42
MyungSik Kim said once upon a time: > This is what I want. If user use terminal mode, they just >connected unix host without asking userid&password (also >unix host doesn't ask userid & password because its hosts.equiv >file has NAS ip address) or ppp then they just be authenticated. >Does anyone implement service like this? You need to setup the accounts via a RADIUS server with Service-Type = Login.
Subject: Re: (usr-tc) PSU in alarm
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-09 12:54:13
> One of my PSU's is in alarm.............I suppose its toast, but is there > anything I should try? > > 1. I have tried reinserting it (powered off to do this of course) > 2. I have checked the one fuse on the PSU > > Any other suggestions? Is this common? Just remember that either the front or the back can fry. I had something blow in the back of one of my power supplies. I wouldn't say it's common, though. That's the only power supply related failure I have had out of the 80 or so chassis I have. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Open code
From: MegaZone <megazone@livingston.com>
Date: 1998-02-09 13:15:18
Once upon a time Pete Ashdown shaped the electrons to say... >The idea of competitors stealing code blows the GNU and the Netscape >model. The GNU license clearly states that you are free to use the code in >any way, just as long as you redistribute the source with it. If And this doesn't get violated everyday? Please... And you don't have to reuse the code as is, you look at it, get a good concept, and implement it your own way. Look ma, I didn't use their code. There is a lot more there than just the lines of code, there are very different ways of doing things overall. >would be forced into redistributing their source as well. Otherwise one >could sue the pants off the other. Don't you think Netscape is thinking Without the source prove that Vendor X swiped from Vendor Y. "Oh, they made PPP better - must have been taken from our code!" C'mon, you think any vendor could risk a bad lawsuit without some real proof? They'd be wide open to major damages if they brought a suit and were proved wrong. >about Microsoft when they're putting out source to Navigator? I've heard a few things about the license, but we'll see. I heard that any changes made remain the domain of Netscape, and that the source is open and free until such time as MS charges for IE. In other words, NN will be free as long as IE is free. But we'll see when the public release is made. But in any case, Netscape isn't doing anything radical. As I said in another post - they make their money off of their server products. And they aren't tossing the source for those out. Netscape releasing the source to Navigator is like Livingston releasing the source to RADIUS, not like releasing ComOS. They also make money directly from ads on their default page, and indirectly via mindshare keeping their product line strong. But financially Navigator was never a cash cow, most people were using it free anyway. >The only reason open code will probably never see the light of day is the >fact that the suits don't understand the concept. They'd rather be None of the engineers I know are keen on the idea either. -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: Re: (usr-tc) Netserver route problem
From: Jeff Lynch <jeff@mercury.jorsm.com>
Date: 1998-02-09 13:53:27
On Mon, 9 Feb 1998, Wayne Barber wrote: > Ummm, Jeff, if you actually read my message you'll see that I'm > asking him not to send two copies to the list. I don't mind the > questions and answers at all (that's why we're all on this list). but > some people like Brian have very similar to: and cc: addresses that > both end up on the list. One is sufficient. > > Thank you, > Wayne Barber - barberw@tidewater.net Internet System Administrator > Coastal Telco Services Sorry Wayne. I've been playing catch-up after being gone for a long weekend and mis-read your message. ========================================================================= Jeffrey A. Lynch, President JORSM Internet email: jeff@jorsm.com Northwest Indiana's Full-Service Provider Voice: (219)322-2180 927 Sheffield Avenue, Dyer, IN 46311 Autoresponse: info@jorsm.com http://www.jorsm.com
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: Ian Roy <iroy@smokies.amnix.com>
Date: 1998-02-09 14:03:21
> Ian Roy <iroy@smokies.amnix.com> writes: > > > How can I set it so only X2 subscribers can use the digital lines and not > > > everyone? > > > > > > Thanks > > > > > > Matthew > (...) > > > > Matthew, > > > > We were looking for a solution to the same problem. We use Merit > > Radius here. > > If you want, you can control this at the modem level without involving > authentication at all. The net result will be a failure to train so > the modem will be tied up for a few seconds while it attempts to > train, but that's really not that much worse than dropping the user > after the fact or during authentication. > > All you need to do is program the modems involved not to work with > modulations other than x2. You can do this by changing some of the > variables in the modem signal control table (which should be > accessible by that name in TCM, or in the mdmScTable table if you are > using the MIB directly). The objects you want to change are: > > Object(s) Set to > ---------------------------------------------------------------- > mdmSc300, mdmSc1200, mdmSc2400 enable > Disables lowest baud rates. The value is actually > enabled since the objects are defined as a negative (if > enabled they prevent the specified connection rate). > > mdmScHstMod disable > Disables USR proprietary HST mode. > > mdmScV32Mod, mdmScV32Bis, mdmScV32BisEnhance, disable > mdmScV32TerboModeEnable > Disables V.32 and variations (9600-19.2K) > > mdmScVFCModeEnable disable > Disables V.FC > > mdmScV34ModeEnable disable > Disables V.34 > > > However, you should ensure that the mdmScV34pModeEnable variable is > enabled (the default), since it comes into play when computing the > back-channel for x2 calls - but having it enabled does not allow V.34+ > connections since that requires the V.34 mode object to be enabled. > > Oh, this assumes modem code 5.x.5 or above (quad, any HDM should be > fine). If for some reason you're running earlier, then you might not > want to lock out the V.34 setting, since that will include even x2 > callers that fall out to V.34. > > This is how we ran (and still run) our x2 "trial" nodes, to restrict > them to x2 callers only. The only kind of a modem that can connect to > hub modems configured with the above objects are those that support > x2. Note that it doesn't require that the final connection be x2 (the > caller may still fall back to V.34 if they don't have an x2 capable > line), only that the calling modem be x2 capable in the first place. > > -- David > > /-----------------------------------------------------------------------\ > \ David Bolen \ Internet: db3l@ans.net / > | ANS Communications, Inc. \ Phone: (914) 701-5327 | > / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ > \-----------------------------------------------------------------------/ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > David, I like your solutions above. But, I have one question: What happens to a user who has not subscribed to X2 services, but has an X2 modem ? Thanks Ian Roy
Subject: Re: (usr-tc) Netserver route problem
From: Wayne Barber <barberw@tidewater.net>
Date: 1998-02-09 14:17:02
> Date: Mon, 9 Feb 1998 11:03:58 -0600 (CST) > From: Jeff Lynch <jeff@mercury.jorsm.com> > On Sun, 8 Feb 1998, Wayne Barber wrote: > > > > > Brian, > > I wish I could help with your problem or answer your questions, but I > > can't. However, would it be possible to get you to stop ccing the > > list? It causes everyone to receive your messages twice (and also > > doubles the replies from some people). > > > I do not concur with this request. Brian is working out a problem that > could very easily help the rest of us when if we have the same > problem. If this was a simple config issue the total solution would > have been posted and completed...as I'm certain will eventually > happen. In that case, I will be happy to see it on this list. > > Dup replies are from people not being careful about who is in the > cc and to lists before sending it. > Ummm, Jeff, if you actually read my message you'll see that I'm asking him not to send two copies to the list. I don't mind the questions and answers at all (that's why we're all on this list). but some people like Brian have very similar to: and cc: addresses that both end up on the list. One is sufficient. Thank you, Wayne Barber - barberw@tidewater.net Internet System Administrator Coastal Telco Services
Subject: Re: (usr-tc) Open code
From: MegaZone <megazone@livingston.com>
Date: 1998-02-09 14:33:22
Once upon a time Pete Ashdown shaped the electrons to say... >I presume that Stallman and other GNU authors don't have the time, >resources, nor the care to pursue this legally. You appear to know very Correct. >little about the GNU license, or why it works. I suggest you read up, You are rather wrong, I am very familiar with it. And I also have heard from peoplr directly "Oh, yeah, I got that idea from g****". So people don't necessarily use the Gnu code, but they profit from it. >What a tragedy. Better code throughout the industry. Come done from the clouds, or share what you're smoking. THIS IS A COMPETITIVE INDUSTRY. Do you understand that Grasshopper? Can you for a moment grasp that these very differences are what feeds thousands of people and keeps them employeed? That the competition is what drives the creative process for most of them to create better ways of doing things - so that they can sell more? If not, you aren't operating in reality. >Real proof is the key. Why would anyone launch a lawsuit over a suspicion? Thanks for proving my point. Vendor X borrows a concept from Vendor Y, but Vendor Y can't prove it. Thuise Vendor Y is screwed. >Besides, whoever gives out the source first would have the advantage. >While other vendors are playing catch-up in copying the latest version, >even better features and optimizations are being added. If one feature were all there were - *maybe*. But what if Vendor X has super feature Fu, and Vendor Y has super feature Bar. Vendor Y gives out their code, and Vendor X improves their version of Bar to match, or maybe surpass, Vendor Y. Now X has A better Fu, and an equal or better Bar - while Vendor Y is still fighting to do the R&D needed to improve their Fu. The competitive advantage is definitely with X here. And being second with a feature is not bad. It has been Livingston's philosophy on many things. Let someone else walk into the minefield first. We have the engineering talent to be able to rapidly develop features and deploy them in a solid form. So watch what X does, go "Ooh, THAT was a mistake" and then avoid it. Many times you can come to market with a feature a little later, but have it stable *before* the other guy because they need to redo things. And redoing takes time. >for their hardware. Once again, all the access-server companies make their >money off of the hardware. The software is incidental. And I already said this statement is completely false. As has already been kindly pointed out. While I did say that the HW and SW for a system, it is true that the software is modular. Most vendors use C for things and write modular functions - not a new concept. We've stated openly that we feel the major strength of the PortMaster product is ComOS - NOT THE HARDWARE. Sure, the HW is good HW, but it isn't anything earth- shaking in design. Few NAS products are. But what let's use use the inexpensive HW and still get killer performance? *COMOS!* Without the extremely tight coding in ComOS a PM would be lame. We make our money from the software - pure and simple. One of the major reasons Lucent targeted Livingston was ComOS. USR licensed ComOS too, recall. Saying the software is incidental shows a complete lack of understanding of the situation. >If you think the complexity of Navigator 4 compares to the complexity of an >authentication server, you really don't have a grip on coding. Besides, You missed the point - not surprising after the others you missed. My point is FOR NETSCAPE Navigator is like RADIUS is for Livingston. And for Netscape, their Server products is like ComOS is for Livignston. Navigator and RADIUS are *enablers* - they help to sell the other products. They add to the feature set, and help to spread the name and reputation of the vendor. But they're not money makers in and of themselves. What they do is help enable the sale of products tht *do* make money. For Netscape that means helping to sell their server products, for Livignston that means helping to sell PortMasters. And that is the same reason for the License on 2.0 and up. Other vendors were using our RADIUS to sell their products. "Well, we don't have a RADIUS server, but you can use Livingston's free." So we stopped allowing that. If you want a free server personally I recommend Cistron RADIUS, but there are plenty to pick from. >we're not talking about the free redistribution of software. We're talking >about giving customers access to the source. The whole reason I use I'll state it clearly: Customers cannot be trusted not to share the source with another vendor. If you think that is extreme, we've already found customers redistributing RADIUS source openly even though the license forbids it. And it is one of the running jokes in the industry that if you tell someone something under NDA it jsut takes a little longer for them to tell someone else. I've had too many people send me things with "I shouldn't be telling you this, but..." If you give the source to anyone, you increase your risk facotor. If you give it to multiple people it is geometric. Just look at the problems Livingston/Lucent and USR/3Com have gone through over the ComOS license. And that is a license between two major vendors! You think anyone is eager to see it repeated with individuals on a mass basis. If you want full access to the source - Linux, GateD, and interface cards. You can, and people do, roll their own access server. You are free to do so, and no one will stop you. You can write your own source to your heart's content and give it away to any and all who ask. You don't HAVE to buy an access server from anyone. We're perfectly willing to let the diehards go and do that - and I suspect all of the major vendors are - because they make up less than 1/10th of 1% of the marketplace. The majority of people don't have the talent, the time, or the desire to build their own solutions. So they look to a vendor to do the work for them. I *could* build my own house, I used to do construction and know how to do framing, wiring, plumbing, etc. But I'd probably hire a builder to do it for me. I'm about to move cross country. I *could* pack a van myself, drive it across country and unload it, but I'm having movers do it. Why? Because I've DONE it myself and it is a big pain in the ass. I'd rather let someone else carry an oak bedroom set around, deal with the drive, and unload it. I could go hunting for my meals - but I prefer to use the Safeway up the street. You pay for the convenience and for expertise that you yourself might not possess. If you can writing everything needed for an access server, from Bellcore signalling to PPP to Stac to BGP - then you really should start your own company. >Merit's RADIUS server rather than Livingstons, or USRs, or Ascends is >because *I* have control over the source, and *I* have made modifications >(which anyone is free to have) that make it work better for *me*. And? You could do the same thing to Livingston's RADIUS - you're free to redistribute diffs at the least. And Merit can always yank the rug out by making major code changes between releases, so you can either split off your tree, or rehack your features. >So far this discussion has generated three customers who exclaim how great >the idea is, and one vendor who pooh-poohs it. What do you do for a living? Run an ISP? Why don't you give away your dialin accounts. Go on, let everyone in for free. Why not give your user information out on the net, let all of your competitors have access. What, doesn't sound like a good idea? I'd love a free, unlimited use account with all the hardrive I can use. -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: Re: (usr-tc) Open code
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 14:56:11
MegaZone said once upon a time: >Once upon a time Pete Ashdown shaped the electrons to say... >>The idea of competitors stealing code blows the GNU and the Netscape >>model. The GNU license clearly states that you are free to use the code in >>any way, just as long as you redistribute the source with it. If > >And this doesn't get violated everyday? Please... I presume that Stallman and other GNU authors don't have the time, resources, nor the care to pursue this legally. You appear to know very little about the GNU license, or why it works. I suggest you read up, starting with: http://www.gnu.org/philosophy/why-free.html >And you don't have to reuse the code as is, you look at it, get a good >concept, and implement it your own way. Look ma, I didn't use their code. >There is a lot more there than just the lines of code, there are very >different ways of doing things overall. What a tragedy. Better code throughout the industry. >Without the source prove that Vendor X swiped from Vendor Y. "Oh, they >made PPP better - must have been taken from our code!" C'mon, you think >any vendor could risk a bad lawsuit without some real proof? They'd be >wide open to major damages if they brought a suit and were proved wrong. Real proof is the key. Why would anyone launch a lawsuit over a suspicion? Besides, whoever gives out the source first would have the advantage. While other vendors are playing catch-up in copying the latest version, even better features and optimizations are being added. >But in any case, Netscape isn't doing anything radical. As I said in >another post - they make their money off of their server products. And >they aren't tossing the source for those out. As I would presume the hardware companies wouldn't put the schematics out for their hardware. Once again, all the access-server companies make their money off of the hardware. The software is incidental. > Netscape releasing the >source to Navigator is like Livingston releasing the source to RADIUS, >not like releasing ComOS. They also make money directly from ads on >their default page, and indirectly via mindshare keeping their product >line strong. But financially Navigator was never a cash cow, most people >were using it free anyway. If you think the complexity of Navigator 4 compares to the complexity of an authentication server, you really don't have a grip on coding. Besides, we're not talking about the free redistribution of software. We're talking about giving customers access to the source. The whole reason I use Merit's RADIUS server rather than Livingstons, or USRs, or Ascends is because *I* have control over the source, and *I* have made modifications (which anyone is free to have) that make it work better for *me*. >>The only reason open code will probably never see the light of day is the >>fact that the suits don't understand the concept. They'd rather be > >None of the engineers I know are keen on the idea either. Probably because most of them would be out of a job. So far this discussion has generated three customers who exclaim how great the idea is, and one vendor who pooh-poohs it.
Subject: Re: (usr-tc) Netserver route problem
From: Russ Panula <rpanula@dacmail.net>
Date: 1998-02-09 15:04:08
On Sat, 7 Feb 1998 21:29:50 -0600 (CST), Brian <signal@shreve.net> wrote: >I have a subnet, declared as follows: > >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > >yet, whenever I start gated on a machine (running rip), I get this in the >routing tables: > >208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 Try adding Framed-Netmask = 255.255.255.255 to your radius user entry. You might also want to add Framed-Routing = None Just to make sure the netserver doesn't listen to any rip routes coming from the connection.
Subject: (usr-tc) Quads going into alarm
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-09 15:08:09
About every 10 days, one of my three TC racks goes into alarm. Every time so far, when I pull up TCM, one of the quad modems is in a yellow state. Generally, the card has calls up, and will accept more calls. The only problem seems to be that the NMC thinks the card is in alarm for some reason. One time, the card actually died when I hardware reset it, but generally, I just hardware reset the card and go on my way. Any ideas so this doesn't continue to happen? Brian
Subject: Re: (usr-tc) PSU in alarm
From: Brian <signal@shreve.net>
Date: 1998-02-09 15:13:13
On Mon, 9 Feb 1998, Curt Shambeau wrote: > > One of my PSU's is in alarm.............I suppose its toast, but is there > > anything I should try? > > > > 1. I have tried reinserting it (powered off to do this of course) > > 2. I have checked the one fuse on the PSU > > > > Any other suggestions? Is this common? > > Just remember that either the front or the back can fry. I had > something blow in the back of one of my power supplies. > > I wouldn't say it's common, though. That's the only power supply related > failure I have had out of the 80 or so chassis I have. Since it was a PSU alarm and not a PSI alarm, I didn't check the back half out. I am assuming its the front half (PSU) thats bad. Brian > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Stuff for sale
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 15:17:27
I've got a bunch of equipment coming up for sale. If anyone is interested in the following please email me directly. Quad v34 Digital Modem NAC $1100 Netserver PRI (20 meg RAM) w/HighSpeed WAN,Ethernet NIC $3250 Dual PRI NIC/NAC $2300 Dual T1 (186) NIC/NAC $2300
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-09 15:19:27
On Mon, 9 Feb 1998, Russ Panula wrote: > On Sat, 7 Feb 1998 21:29:50 -0600 (CST), Brian <signal@shreve.net> > wrote: > > >I have a subnet, declared as follows: > > > >signal Authentication-Type = Unix-PW > > Service-Type = Framed, > > Framed-Protocol = PPP, > > Framed-IP-Address = 208.214.45.1, > > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > > Framed-MTU = 1500, > > Framed-Compression = Van-Jacobson-TCP-IP > > > >yet, whenever I start gated on a machine (running rip), I get this in the > >routing tables: > > > >208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 > > Try adding > > Framed-Netmask = 255.255.255.255 Why would I delcare a Framed-Netmask on a Framed-Route account? > > to your radius user entry. You might also want to add > > Framed-Routing = None This has the reverse effect when running Merit RADIUS, which I run. Adding "Framed-Routing = None" actually enables rip updates, whos fault this is, I dont know, but its an anomaly between USR hubs and Merit RADIUS which can be verified by doing a show SXX on the netserver and you will see "RIPsend" > > Just to make sure the netserver doesn't listen to any rip routes > coming from the connection. Brian > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: Charles Hill <chill@ionet.net>
Date: 1998-02-09 15:41:45
On Mon, 9 Feb 1998, Ian Roy wrote: > David, > > I like your solutions above. But, I have one question: > > What happens to a user who has not subscribed to X2 services, but > has an X2 modem ? > > Thanks > > Ian Roy If you disable x2 in the modems and want to prevent non-x2 subscribers from using the x2-enabled NAS, you'll probably have to hack up your radius so that non-x2 subscribers can't authenticate on that particular NAS. -CH
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: David Bolen <db3l@ans.net>
Date: 1998-02-09 15:53:34
Ian Roy <iroy@smokies.amnix.com> writes: > > How can I set it so only X2 subscribers can use the digital lines and not > > everyone? > > > > Thanks > > > > Matthew (...) > > Matthew, > > We were looking for a solution to the same problem. We use Merit > Radius here. If you want, you can control this at the modem level without involving authentication at all. The net result will be a failure to train so the modem will be tied up for a few seconds while it attempts to train, but that's really not that much worse than dropping the user after the fact or during authentication. All you need to do is program the modems involved not to work with modulations other than x2. You can do this by changing some of the variables in the modem signal control table (which should be accessible by that name in TCM, or in the mdmScTable table if you are using the MIB directly). The objects you want to change are: Object(s) Set to ---------------------------------------------------------------- mdmSc300, mdmSc1200, mdmSc2400 enable Disables lowest baud rates. The value is actually enabled since the objects are defined as a negative (if enabled they prevent the specified connection rate). mdmScHstMod disable Disables USR proprietary HST mode. mdmScV32Mod, mdmScV32Bis, mdmScV32BisEnhance, disable mdmScV32TerboModeEnable Disables V.32 and variations (9600-19.2K) mdmScVFCModeEnable disable Disables V.FC mdmScV34ModeEnable disable Disables V.34 However, you should ensure that the mdmScV34pModeEnable variable is enabled (the default), since it comes into play when computing the back-channel for x2 calls - but having it enabled does not allow V.34+ connections since that requires the V.34 mode object to be enabled. Oh, this assumes modem code 5.x.5 or above (quad, any HDM should be fine). If for some reason you're running earlier, then you might not want to lock out the V.34 setting, since that will include even x2 callers that fall out to V.34. This is how we ran (and still run) our x2 "trial" nodes, to restrict them to x2 callers only. The only kind of a modem that can connect to hub modems configured with the above objects are those that support x2. Note that it doesn't require that the final connection be x2 (the caller may still fall back to V.34 if they don't have an x2 capable line), only that the calling modem be x2 capable in the first place. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) livingston radius non-functional
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-09 16:09:34
On Mon, 9 Feb 1998, Jay M Christner wrote: > Hmmmm, I was told (by someone AT LIVINGSTON) that it was okay to use the > server (RadiusNT 2.0 was in question at the time) with other equipment, but > it was completely unsupported. Seems sorta contradictory. Anyway, it's a > moot point for me now, but there seems to be some dissention in the ranks. :) You can use Lucent radius with other NASes just fine. You can only use Lucent radius 2.0 or higher if you own a piece of Lucent RABU gear. Brian
Subject: Re: (usr-tc) Netserver route problem
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-09 16:11:23
At 03:04 PM 2/9/98 -0600, Russ Panula wrote: > Framed-Routing = None > >Just to make sure the netserver doesn't listen to any rip routes >coming from the connection. No actually this is a bad idea. The best idea is to not define that variable all together and set the default routing on the netserver to none.. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) Open code
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-09 16:13:12
MegaZone said once upon a time: >Come done from the clouds, or share what you're smoking. THIS IS A >COMPETITIVE INDUSTRY. Do you understand that Grasshopper? Can you for >a moment grasp that these very differences are what feeds thousands of >people and keeps them employeed? Can you, for a single moment understand that customers are pissed throughout this competitive industry? That choosing an access server is like choosing the lesser of several evils? That the software for access servers and dealing with the bugs is a key problem that nobody seems to know a good work around? The problem with arguing this point with you is the fact that you think your own product can do no wrong. >Few NAS products are. But what let's use use the >inexpensive HW and still get killer performance? *COMOS!* Without the >extremely tight coding in ComOS a PM would be lame. We make our money >from the software - pure and simple. One of the major reasons Lucent >targeted Livingston was ComOS. USR licensed ComOS too, recall. I, for one, am extremely happy they dumped that code base. >If you think that is extreme, we've already found customers redistributing >RADIUS source openly even though the license forbids it. Oh God! What a tragedy! Burn the offices! You guys are going down! >If you want full access to the source - Linux, GateD, and interface cards. >You can, and people do, roll their own access server. You are free to do >so, and no one will stop you. You can write your own source to your heart's >content and give it away to any and all who ask. You don't HAVE to buy an >access server from anyone. Amazingly enough, most of the hardware around here runs that way, and I must say it runs better than anything I don't have the source to. If someone made a PRI/X2 interface card for the PC, I would most likely build my own access server. >We're perfectly willing to let the diehards go and do that - and I suspect >all of the major vendors are - because they make up less than 1/10th of 1% >of the marketplace. The majority of people don't have the talent, the time, >or the desire to build their own solutions. So they look to a vendor to >do the work for them. Your silent majority generally doesn't know the difference between a subnet and a phone number. Unfortunately, vendors take one month to never to resolve issues with their code. Issues that usually take an hour to fix once it lands on the desk of someone who can actually touch the tree. [Much ranting deleted] >I could go hunting for my meals - but I prefer to use the Safeway up the >street. [Much ranting deleted] >You pay for the convenience and for expertise that you yourself might not >possess. If you can writing everything needed for an access server, from >Bellcore signalling to PPP to Stac to BGP - then you really should start >your own company. I'm already running one, and yes, I do understand most of the complexities involved, from the wire to the screen. I'm not stating that I could write better code, but I'm saying that I'm happy to contribute, for free. >And? You could do the same thing to Livingston's RADIUS - you're free to >redistribute diffs at the least. And Merit can always yank the rug out >by making major code changes between releases, so you can either split off >your tree, or rehack your features. This is what tools are for. >What do you do for a living? Run an ISP? Why don't you give away your >dialin accounts. Go on, let everyone in for free. Why not give your >user information out on the net, let all of your competitors have access. > >What, doesn't sound like a good idea? I'd love a free, unlimited use >account with all the hardrive I can use. Tell you what. Invest in the hardware to support the account and you can have one. This is an incredibly lame comparison. Spend what I spend on Cisco and USR and be satisfied with "no" and long waits for reasonably simple software fixes and you'll understand why I'd like to be able to peek at the source to fix my own problems. Make me sign away my life, I'll give my first born if I give the code to anyone. Just don't make me wait two months to get a conditional added. The core of this issue is that the customer suffers in order to protect the company's property. Customer suffering? TOO BAD, WE'RE HERE TO MAKE MONEY! Someone isn't putting two and two together.
Subject: Re: (usr-tc) Open code
From: MegaZone <megazone@livingston.com>
Date: 1998-02-09 16:19:33
Once upon a time Pete Ashdown shaped the electrons to say... >Can you, for a single moment understand that customers are pissed >throughout this competitive industry? That choosing an access server is Frankly, no. Since the very much overwhelming majority of customers I've seen are frankly quite happy. Maybe that's because I deal mostly with our customers - who knows. There are always people upset by some bug, and there are always bugs. But rarely does that become a significant proportion of the total userbase. And there are also always people who just like to be upset and will complain just to complain. >know a good work around? The problem with arguing this point with you is >the fact that you think your own product can do no wrong. Bullshit. How many times over the years have I *personally* announced bug reports and acknowlegements to the user base? I did so just a couple of days ago. Do I think the PortMaster is the best remote access product on the market? Unquestionably, yes, I do. Do I think it can do everything for everyone? No - and I have repeatedly recommended other vendors to customers who needed specific features that we do not support. >I, for one, am extremely happy they dumped that code base. You should check your facts... >must say it runs better than anything I don't have the source to. If >someone made a PRI/X2 interface card for the PC, I would most likely build >my own access server. I believe there are CT1/PRI & K56flex cards for the PC, one would expect they'll support V.90 shortly. Rumage... <URL:http://www.ariel.com/> Designed for use with NT, but you could probably create drivers for Linux (if they haven't been done already). rumage... ---cut--- 5. Does the T1-Modem+ support Linux drivers? No. T1-Modem+ currently only supports UnixWare 2.0 and Windows NT 4.0 drivers. A Device Driver Development Kit (DDK) for custom and embedded applications is available. ---cut--- So you can get the DDK - I'm sure you'd be popular with the Linux community if you got this working. AFAIK X2 servers are still 3Com and Bay Networks only though. >and a phone number. Unfortunately, vendors take one month to never to >resolve issues with their code. Issues that usually take an hour to fix >once it lands on the desk of someone who can actually touch the tree. Most of the bugs we get reported are fixed within days, a few I've seen fixed within hours. Obviously cosmetic issues don't get a rush, but network harming issues go straight to top priority over everything else. Nearly all of the 'long term' problems have come down to modem issues, which originally we didn't have much say in. Being part of the same company has helped that to some degree, as there is a more directly path between engineering groups. There is also a difference in bug reports. Saying "I think there is a bug" then means we have to track it down - I remember one case that took weeks to track down, but minutes to fix. Once it was found it wasn't hard at all - it was just hard to find. And those are the ones that tend to get harped on for a long time by a few vocal people. As far as we're concerned fixing bugs should come first - it is a problem in an existing implementation, so it comes before adding new things to the code. I can't say that all vendors are the same, apparently not. >The core of this issue is that the customer suffers in order to protect the >company's property. Customer suffering? TOO BAD, WE'RE HERE TO MAKE >MONEY! Someone isn't putting two and two together. So shop elsewhere or build it yourself. Vote with your wallet. If you don't like something, don't buy it - buy something else or do it yourself. -MZ -- Lucent Remote Access Division - Chair, Department of Interstitial Affairs Phone: 800-458-9966 510-737-2100 FAX: 510-737-2110 megazone@livingston.com For support requests: support@livingston.com <http://www.livingston.com/> Snail mail: 4464 Willow Road, Pleasanton, CA 94588
Subject: (usr-tc) Netserver Config Example
From: Brian <signal@shreve.net>
Date: 1998-02-09 16:34:31
Here is an example of how to configure a netserver, well commented, since I saw some requests for something like this in the past. This netserver has 12 Quad modems, and no HDM's. It acts as an MPIP server for 4 Hubs. I thought I would throw this out there to not only help anyone who needs a config commented, but also in hopes of getting comments and critisizm about how I am doing things. If you see I have left something out let me know. I would be interested in seeing others HiPer ARC configs, commented. I have Pete's which is great, but the more I see how others are doing things, the better I can tune my own. All my netservers are on a network seperate from there pools. Netservers: 208.206.76.0/24 IP Pools are /26's set sysname usr1ts1 # name of the netserver set net0 address 208.206.76.35 # ip of the netserver set net0 netmask 255.255.255.0 # netmask of the netserver set gateway 208.206.76.1 # gateway of the netserver set domain shreve.net # our domain set nameserver 1 208.206.76.2 # primary nameserver set nameserver 2 208.206.76.3 # secondary nameserver set authentic 208.206.76.5 # RADIUS server ip set alternate 208.206.76.23 # secondary RADIUS server ip set secret secret # RADIUS shared secret set accounting 208.206.76.5 # RADIUS accounting server ip set assigned 208.214.44.1 # First usable IP to assign set limit 46 # number of B channels in this box set modem s5-s52 active # first and last modem set pppmodem on # PPP comes from the modem set all login network dialin # usable for PPP or terminal set all security on # Don't let people access modems add filter guest.in # Our guest filter set filter guest.in 1 deny 0.0.0.0/0 0.0.0.0/0 tcp dst lt 24 set filter guest.in 2 deny 0.0.0.0/0 0.0.0.0/0 tcp dst eq 110 set filter guest.in 3 deny 0.0.0.0/0 0.0.0.0/0 tcp dst eq 25 set filter guest.in 4 permit add filter guest.out set filter guest.out 1 deny 0.0.0.0/0 0.0.0.0/0 tcp src lt 24 set filter guest.out 2 deny 0.0.0.0/0 0.0.0.0/0 tcp src eq 110 set filter guest.out 3 deny 0.0.0.0/0 0.0.0.0/0 tcp src eq 25 set filter guest.out 4 permit add filter mail.in # filter for mail only accounts set filter mail.in 1 permit 0.0.0.0/0 208.206.76.2/32 tcp dst eq 106 set filter mail.in 2 permit 0.0.0.0/0 208.206.76.5/32 tcp dst eq 106 set filter mail.in 3 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 106 set filter mail.in 4 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 109 set filter mail.in 5 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 110 set filter mail.in 6 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 143 set filter mail.in 7 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 25 set filter mail.in 8 permit 0.0.0.0/0 208.206.76.2/32 udp dst eq 53 set filter mail.in 9 permit 0.0.0.0/0 208.206.76.3/32 udp dst eq 53 set filter mail.in 10 permit 0.0.0.0/0 208.206.76.2/32 tcp dst eq 53 set filter mail.in 11 permit 0.0.0.0/0 208.206.76.3/32 tcp dst eq 53 set filter mail.in 12 permit 0.0.0.0/0 208.206.76.5/24 tcp dst eq 80 set filter mail.in 13 permit 0.0.0.0/0 208.206.76.5/24 tcp dst eq 8000 set filter mail.in 14 permit 0.0.0.0/0 208.206.76.5/24 tcp dst eq 443 set filter mail.in 15 deny 0.0.0.0/0 0.0.0.0/0 add filter mail.out set filter mail.out 1 permit 208.206.76.0/24 0.0.0.0/0 tcp estab set filter mail.out 2 permit 208.206.76.2/32 0.0.0.0/0 udp src eq 53 set filter mail.out 3 permit 208.206.76.3/32 0.0.0.0/0 udp src eq 53 set filter mail.out 4 deny 0.0.0.0/0 0.0.0.0/0 set proxyarp on # proxyarp the IP address set lanwan_routing off # don't route between lan and wan set snmp off # no snmp on netserver set net0 routing on # broadcast and listen to RIP set ripv2 on # use RIPv2 set enh_routing on # used enhanced routing add ippool dialup1 208.214.44.1 62 # to aggregate routes add netmask 208.214.44.0 255.255.255.192 set password secret # the !root password set time 1 192.43.244.18 # the time server IP set mpipserver on # enable MPIP server set mpipserver 1 208.206.76.35 secret # use itself as MPIP server add mpipclient 208.206.76.35 secret # add MPIP clients add mpipclient 208.206.76.37 secret add mpipclient 208.206.76.39 secret add mpipclient 208.206.76.41 secret save all # Save to NVRAM reboot # reboot /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: David Bolen <db3l@ans.net>
Date: 1998-02-09 16:35:30
Ian Roy <iroy@smokies.amnix.com> writes: > I like your solutions above. But, I have one question: > > What happens to a user who has not subscribed to X2 services, but > has an X2 modem ? Assuming that the phrase "X2 services" refers to some sort of service model or access plan that a user must sign up for specifically, then obviously the modems can't tell - as long as the user has an x2 capable client modem they'll be allowed to train successfully by the server modems. I may have misinterpreted "X2 subscribers" in Matthew's original request to mean "subscribers who have an x2 modem" rather than users who subscribe to a separate x2-supporting service. In the former case - where you offer x2 services to your subscribers who have an appropriate modem but don't want non-x2 modems to tie up your better quality lines - controlling it at the modem should work fine. But in the latter case - where you need to know if the user is paying for x2 service - it's not sufficient and you will need to do something at some other layer such as authentication. I suppose the modem restrictions might be marginally helpful even in the separate service case in terms of throttling back on the number of requests you need to work on and/or helping to get rid of non-x2 people faster than authentication or a periodic scan might be able to. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Open code
From: David Bolen <db3l@ans.net>
Date: 1998-02-09 16:43:43
MegaZone <megazone@livingston.com> writes: > >The only reason open code will probably never see the light of day is the > >fact that the suits don't understand the concept. They'd rather be > > None of the engineers I know are keen on the idea either. I have to admit to being pretty split on the idea myself. Part of me likes to distribute source code for everything (although to be honest, the GPL grates on me at times in the way that it "infects" anything it touches so I'm not positive that's the best model, but that's a separate discussion) and would love to see it done. However, part of me can't quite see it happening, or perhaps being even reasonable. Much of what makes each of these systems what they are is in the software - that's where the competitive advantage (or disadvantage) lies, and where much of the intellectual property is. There have been some comments made on the list about such source being hardware specific, and for a small portion of it I'm sure that's true. But without knowing specifics, I'd also bet money that any of the major vendors source is, at this point, modular enough that such hardware interfacing is in device drivers, or otherwise low level code that is easily replaced. I'd also bet that shortly after such source were released, it would be easy to just take out those pieces, replace it with something else (say to run on your average hardware box with multi-port serial interfaces) and avoid buying the hardware (or further software) entirely. It's a slippery slope that I find it hard to believe any of the companies would want to embark on. We can always wish though... -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: (usr-tc) Pipelines and Netservers
From: Brian <signal@shreve.net>
Date: 1998-02-09 17:05:42
The last time I checked, you can't tell a Pipeline "Use the next hop as the gateway address", instead you *have* to tell it a default route. Well I am sure the Pipelines have always played nicely with Ascend NAS equipment. How are you all getting around this? I just point the Pipelines default route at our border router, and on a Netserver equipted chassis that works. Some of you if I remember right, use the "set reported" function. Is this the way to do it? Say I have 3 hubs: 208.206.76.35 208.206.76.37 208.206.76.39 Do I just take an available unused IP like, 208.206.76.45, and do this on all 3 hubs: set reported 208.206.76.45 and then set the pipelines default routes at it? Secondly, is there an equivelent command on the ARC, to help Pipelines along.........similar to a "set reported". What it boils down to, is our Pipelines are configured like I told you, pointing at our border router, which totally bites, I know. Perhaps now Pipelines will play nice and accept the syntax: Destination: 0.0.0.0 Gateway: 0.0.0.0 What happens, is with the Netservers, it actually works (pointing Pipelines at our border router), albeit probably inefficient as hell. But on the ARC, we have some Pipeline customers complaining that it will freeze up sometimes, disallowing various services, while still allowing others, basically total routing instability. The Pipeline is a very popular ISDN router, and we are getting alot of customers using them. So i am looking for the ideal way to have the pipeline talk to our hubs, whether that means using "set reported" on the Netservers, and something else on the ARC's or whatever. Appreciate any help, Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quads going into alarm
From: Brian <signal@shreve.net>
Date: 1998-02-09 17:08:59
On Mon, 9 Feb 1998, Brian Elfert wrote: > About every 10 days, one of my three TC racks goes into alarm. > > Every time so far, when I pull up TCM, one of the quad modems is in a > yellow state. Generally, the card has calls up, and will accept more > calls. The only problem seems to be that the NMC thinks the card is in > alarm for some reason. > > One time, the card actually died when I hardware reset it, but generally, > I just hardware reset the card and go on my way. > > Any ideas so this doesn't continue to happen? > > Brian make sure you have done a "save chassis to nvram" on the nmc card. Brian > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Address pool for portion of TC
From: Beth Montes <bmontes@iland.net>
Date: 1998-02-09 17:27:30
If I want to use an address pool for most, but not all, of my USR TC, and want to assign specific IP's to the remaining ports, how do I accomplish that? I tried just setting the required IP on the specific port(s), using 'set s# address'. A 'show s#' shows Active Configuration Default Configuration -------------------- --------------------- Remote Host: 0.0.0.0 205.242.230.202 When a dial-in user connects, an address from the pool still gets assigned, and my specified address is ignored. Beth _____________________________________________ Beth Montes bmontes@iland.net 816-827-5111 x21 ComputerLand of Sedalia http://www.c-land.com I-Land Internet Services http://www.iland.net _____________________________________________
Subject: Re: (usr-tc) Netserver Config Example
From: Jeremy Lingmann <jeremy@europa.com>
Date: 1998-02-09 18:00:21
Brian, Some questions on the following commands: > set enh_routing on # used enhanced routing > add ippool dialup1 208.214.44.1 62 # to aggregate routes Are you saying that by enabling both enh_routing and defining an ippool the Total Control will then aggregate it's RIP route announcements? I've been trying to find a way to do this for quite some time. However, from the documentation in the release notes on these two commands, I don't see anything pertaining to that... Any comments? (Just for clarification, I'm thinking of the word "aggregate" in terms of the netserver announcing all of it's assigned IP addresses through a single *ie* /26 RIP announcement, instead of a million individual /32 announcements.) -jeremy On Mon, 9 Feb 1998, Brian wrote: > Here is an example of how to configure a netserver, well commented, since > I saw some requests for something like this in the past. > > This netserver has 12 Quad modems, and no HDM's. It acts as an MPIP > server for 4 Hubs. I thought I would throw this out there to not only > help anyone who needs a config commented, but also in hopes of getting > comments and critisizm about how I am doing things. If you see I have > left something out let me know. > > I would be interested in seeing others HiPer ARC configs, commented. I > have Pete's which is great, but the more I see how others are doing > things, the better I can tune my own. > > All my netservers are on a network seperate from there pools. > > Netservers: 208.206.76.0/24 > IP Pools are /26's > > set sysname usr1ts1 # name of the netserver > set net0 address 208.206.76.35 # ip of the netserver > set net0 netmask 255.255.255.0 # netmask of the netserver > set gateway 208.206.76.1 # gateway of the netserver > set domain shreve.net # our domain > set nameserver 1 208.206.76.2 # primary nameserver > set nameserver 2 208.206.76.3 # secondary nameserver > set authentic 208.206.76.5 # RADIUS server ip > set alternate 208.206.76.23 # secondary RADIUS server ip > set secret secret # RADIUS shared secret > set accounting 208.206.76.5 # RADIUS accounting server ip > set assigned 208.214.44.1 # First usable IP to assign > set limit 46 # number of B channels in this box > set modem s5-s52 active # first and last modem > set pppmodem on # PPP comes from the modem > set all login network dialin # usable for PPP or terminal > set all security on # Don't let people access modems > add filter guest.in # Our guest filter > set filter guest.in 1 deny 0.0.0.0/0 0.0.0.0/0 tcp dst lt 24 > set filter guest.in 2 deny 0.0.0.0/0 0.0.0.0/0 tcp dst eq 110 > set filter guest.in 3 deny 0.0.0.0/0 0.0.0.0/0 tcp dst eq 25 > set filter guest.in 4 permit > add filter guest.out > set filter guest.out 1 deny 0.0.0.0/0 0.0.0.0/0 tcp src lt 24 > set filter guest.out 2 deny 0.0.0.0/0 0.0.0.0/0 tcp src eq 110 > set filter guest.out 3 deny 0.0.0.0/0 0.0.0.0/0 tcp src eq 25 > set filter guest.out 4 permit > add filter mail.in # filter for mail only accounts > set filter mail.in 1 permit 0.0.0.0/0 208.206.76.2/32 tcp dst eq 106 > set filter mail.in 2 permit 0.0.0.0/0 208.206.76.5/32 tcp dst eq 106 > set filter mail.in 3 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 106 > set filter mail.in 4 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 109 > set filter mail.in 5 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 110 > set filter mail.in 6 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 143 > set filter mail.in 7 permit 0.0.0.0/0 208.206.76.23/32 tcp dst eq 25 > set filter mail.in 8 permit 0.0.0.0/0 208.206.76.2/32 udp dst eq 53 > set filter mail.in 9 permit 0.0.0.0/0 208.206.76.3/32 udp dst eq 53 > set filter mail.in 10 permit 0.0.0.0/0 208.206.76.2/32 tcp dst eq 53 > set filter mail.in 11 permit 0.0.0.0/0 208.206.76.3/32 tcp dst eq 53 > set filter mail.in 12 permit 0.0.0.0/0 208.206.76.5/24 tcp dst eq 80 > set filter mail.in 13 permit 0.0.0.0/0 208.206.76.5/24 tcp dst eq 8000 > set filter mail.in 14 permit 0.0.0.0/0 208.206.76.5/24 tcp dst eq 443 > set filter mail.in 15 deny 0.0.0.0/0 0.0.0.0/0 > add filter mail.out > set filter mail.out 1 permit 208.206.76.0/24 0.0.0.0/0 tcp estab > set filter mail.out 2 permit 208.206.76.2/32 0.0.0.0/0 udp src eq 53 > set filter mail.out 3 permit 208.206.76.3/32 0.0.0.0/0 udp src eq 53 > set filter mail.out 4 deny 0.0.0.0/0 0.0.0.0/0 > set proxyarp on # proxyarp the IP address > set lanwan_routing off # don't route between lan and wan > set snmp off # no snmp on netserver > set net0 routing on # broadcast and listen to RIP > set ripv2 on # use RIPv2 > set enh_routing on # used enhanced routing > add ippool dialup1 208.214.44.1 62 # to aggregate routes > add netmask 208.214.44.0 255.255.255.192 > set password secret # the !root password > set time 1 192.43.244.18 # the time server IP > set mpipserver on # enable MPIP server > set mpipserver 1 208.206.76.35 secret # use itself as MPIP server > add mpipclient 208.206.76.35 secret # add MPIP clients > add mpipclient 208.206.76.37 secret > add mpipclient 208.206.76.39 secret > add mpipclient 208.206.76.41 secret > save all # Save to NVRAM > reboot # reboot > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ -- .------------------------------------------------------------------. | Jeremy Lingmann \ http://www | jeremy@europa.com | | Network Manager \ .europa | fax: 503-796-9134 | | Europa Communications, Inc. \ .com | phone: 503-222-9508 |
Subject: (usr-tc) MPIP & ISDN call termination
From: Jeremy Lingmann <jeremy@europa.com>
Date: 1998-02-09 18:34:18
I have two fairly quick (and technical) questions concerning the Total Control system.... First of all, does Netserver MPIP (multi-chassis multilink ppp) work when ISDN calls are terminated on the latest version of the Quad I-Modem cards? We are currently terminating all of our ISDN calls on Netserver cards (7 chassis connected to a total of 14 PRI's) and for performance/monitoring reasons we'd like to change it so that the Modem cards take over the load. If this isn't supported, is it being worked on at all? Secondly, in a hunt group of this size what is the ideal number of mpipserver's to run? We currently have one, and I'd like to set up one or two of the chassis' as backup - but not if there are any possible complications/probems. Does anyone have experience with this? Considering how finicky MPIP can be, it scares me to think of the possibilities... ;-) Any help would be greatly appreciated. It's really hard for us to experiment with these kinds of things in a production environment (of limited funds <sigh>). Thanks, -jeremy -- .------------------------------------------------------------------. | Jeremy Lingmann \ http://www | jeremy@europa.com | | Network Manager \ .europa | fax: 503-796-9134 | | Europa Communications, Inc. \ .com | phone: 503-222-9508 |
Subject: Re: (usr-tc) Open code
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-09 19:25:04
At 12:51 PM 2/9/98 -0700, Pete Ashdown wrote: >The only reason open code will probably never see the light of day is the >fact that the suits don't understand the concept. Maybe, maybe not.. Suits like projections.. No, they love them. If they project even .01% more sales from one method over another, that's what they choose. And most projections are short term no matter what the forcaster may say.. In the short term, proprietary looks better on paper. Alot better.. It's "old hat" and it's safe. That's why suits all look/act alike.. I hate that. As a one time engineer, you won't catch me in no monkey suit! Nike's, Levi's and a patagonia t-shirt for me. Open code, in contrast, can be revolutionary in scope over the long term. Can change the way we do things, etc. What's that 2 dollar word? Paradigm.. That's pretty scarry for someone who just wants to look, act, and get paid the same as all the other suits... the bigger a company gets, the worst this becomes.. until they no longer have any creativity left.. only pro- jections.. *lots* of these.. >They'd rather be >protectionary and hostile towards any outside eyes friendly or not. In the >process they'll be dumping zillions on development and support. > True. Perhaps someone here on this list might have shead light on the quake lag issue. Or maybe implemented OSPF or MPIP in the arc. Just like many of you save USR support dollars.. It's just too hard to "model"... Their projections/analysis just can't cover areas like this. They probably think (from their projections) that their support budget is what supports their equipement.. heheh.. am _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Open code
From: Mark R. Levinson <mrl@isc.upenn.edu>
Date: 1998-02-09 19:27:50
pashdown@xmission.com (Pete Ashdown) writes: >With Netscape giving out their code, the question I have is, why don't >hardware companies do the same? I've already spent the money on the >hardware and the software is for all intent purposes, free. Personally, I would be thrilled to have source code available even if it weren't free. You can get operating system source code from many of the commercial UNIX vendors for a fee (and without the right to redistribute their sources). For some of us it's well worth the money. If Sun et al. can sell source code licenses to end-users without jeopardizing their competitive advantages, it's not obvious why USR couldn't do the same. -Mark- -- Mark R. Levinson, Network Engineer Information Systems and Computing University of Pennsylvania mrl@isc.upenn.edu
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-09 20:17:48
At 03:41 PM 2/9/98 -0600, Charles Hill wrote: >On Mon, 9 Feb 1998, Ian Roy wrote: > >> David, >> >> I like your solutions above. But, I have one question: >> >> What happens to a user who has not subscribed to X2 services, but >> has an X2 modem ? >> >> Thanks >> >> Ian Roy > > >If you disable x2 in the modems and want to prevent non-x2 subscribers >from using the x2-enabled NAS, you'll probably have to hack up your radius >so that non-x2 subscribers can't authenticate on that particular NAS. > >-CH Why waste good x2 modems/digital lines?.. Analog lines cost around 20% less in these parts.. unless you just have to go with one huntgroup, it seems like a good way to disable x2.. But we are fully digital and charge the same for 33.6, 56k, and isdn (1B+D).. Keeps things easy.. am
Subject: (usr-tc) Call Me Stupid
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-09 20:31:00
Ok folks. I thought I could understand just about anything but after a conversation with a 3Com representative, my head is spinning and I am totally confused. Ok, here's the situation. Last year we bought one of the ISP bundles from USR, 1706 I believe, which included the Total Control Chassis, TCM software and the Windows based security and accounting server software (i.e. RADIUS). After 90 days, I lost access to all of the updates on the TotalService website (with no prior notice mind you) and a quick call to now 3Com, I was told we needed a support agreement. Ok, next quick call to Source Technology and I am told we need the PCAP agreement, so we shelled out $1800+ and got the highest agreement they had. Up until now I was under the impression we were all set. Now January comes along and the new RADIUS software gets posted on the Total Service website. A call to the support line sends me to someone else at 3Com who handles access to the site. Now comes the confusion. In talking to a new lady there she attempts to explain to me that I don't have what I need to get RADIUS updates. So I attempt to ask her why ? She says the ISP bundles "run out" after a certain amount of time. I tried to get her to explain what she means. She then goes on to say the RADIUS software isn't included under the CAP agreement and what I really need is part number 001611-00, which ends up being the RADIUS software. Now, in my obvious confusion, I ask her why I should buy it again, since it was included in the bundle, which in my mind is nothing more than a sale price instead of me buying the parts individually. She then again says that it doesn't work this way. So I ask her if I buy the RADIUS software again for the nearly $2K list (according to her, I haven't verified this price yet), what happens at the end of another year ? Is there a software maintenance agreement I need to get. Again, she says "No". At the end of the year I need to purchase the software again ??? Now, call me stupid but I have never heard of such a thing in all of my years buying software. I attempted to explain to her that this was nothing short of highway robbery and they have many competitors selling RADIUS software at 1/2 the price with yearly maintenance costs at a fraction of what she was saying. So my question to anyone else using 3Com's RADIUS software; is this true ? I am going to call Source Technology tomorrow to discuss this. She did mention that many otehrs have called her with the same confusion. Based upon what she is saying, I can see why. I also fail to see why someone would pay $1800+ for full hardware maintenance and phone support and then pay another $2K for new software. So am I out in left field on this ? Jeff Binkley ASA Network Computing
Subject: (usr-tc) Call Me Stupid
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-09 20:31:00
Ok folks. I thought I could understand just about anything but after a conversation with a 3Com representative, my head is spinning and I am totally confused. Ok, here's the situation. Last year we bought one of the ISP bundles from USR, 1706 I believe, which included the Total Control Chassis, TCM software and the Windows based security and accounting server software (i.e. RADIUS). After 90 days, I lost access to all of the updates on the TotalService website (with no prior notice mind you) and a quick call to now 3Com, I was told we needed a support agreement. Ok, next quick call to Source Technology and I am told we need the PCAP agreement, so we shelled out $1800+ and got the highest agreement they had. Up until now I was under the impression we were all set. Now January comes along and the new RADIUS software gets posted on the Total Service website. A call to the support line sends me to someone else at 3Com who handles access to the site. Now comes the confusion. In talking to a new lady there she attempts to explain to me that I don't have what I need to get RADIUS updates. So I attempt to ask her why ? She says the ISP bundles "run out" after a certain amount of time. I tried to get her to explain what she means. She then goes on to say the RADIUS software isn't included under the CAP agreement and what I really need is part number 001611-00, which ends up being the RADIUS software. Now, in my obvious confusion, I ask her why I should buy it again, since it was included in the bundle, which in my mind is nothing more than a sale price instead of me buying the parts individually. She then again says that it doesn't work this way. So I ask her if I buy the RADIUS software again for the nearly $2K list (according to her, I haven't verified this price yet), what happens at the end of another year ? Is there a software maintenance agreement I need to get. Again, she says "No". At the end of the year I need to purchase the software again ??? Now, call me stupid but I have never heard of such a thing in all of my years buying software. I attempted to explain to her that this was nothing short of highway robbery and they have many competitors selling RADIUS software at 1/2 the price with yearly maintenance costs at a fraction of what she was saying. So my question to anyone else using 3Com's RADIUS software; is this true ? I am going to call Source Technology tomorrow to discuss this. She did mention that many otehrs have called her with the same confusion. Based upon what she is saying, I can see why. I also fail to see why someone would pay $1800+ for full hardware maintenance and phone support and then pay another $2K for new software. So am I out in left field on this ? Jeff Binkley ASA Network Computing
Subject: (usr-tc) Call Me Stupid
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-09 20:31:00
Ok folks. I thought I could understand just about anything but after a conversation with a 3Com representative, my head is spinning and I am totally confused. Ok, here's the situation. Last year we bought one of the ISP bundles from USR, 1706 I believe, which included the Total Control Chassis, TCM software and the Windows based security and accounting server software (i.e. RADIUS). After 90 days, I lost access to all of the updates on the TotalService website (with no prior notice mind you) and a quick call to now 3Com, I was told we needed a support agreement. Ok, next quick call to Source Technology and I am told we need the PCAP agreement, so we shelled out $1800+ and got the highest agreement they had. Up until now I was under the impression we were all set. Now January comes along and the new RADIUS software gets posted on the Total Service website. A call to the support line sends me to someone else at 3Com who handles access to the site. Now comes the confusion. In talking to a new lady there she attempts to explain to me that I don't have what I need to get RADIUS updates. So I attempt to ask her why ? She says the ISP bundles "run out" after a certain amount of time. I tried to get her to explain what she means. She then goes on to say the RADIUS software isn't included under the CAP agreement and what I really need is part number 001611-00, which ends up being the RADIUS software. Now, in my obvious confusion, I ask her why I should buy it again, since it was included in the bundle, which in my mind is nothing more than a sale price instead of me buying the parts individually. She then again says that it doesn't work this way. So I ask her if I buy the RADIUS software again for the nearly $2K list (according to her, I haven't verified this price yet), what happens at the end of another year ? Is there a software maintenance agreement I need to get. Again, she says "No". At the end of the year I need to purchase the software again ??? Now, call me stupid but I have never heard of such a thing in all of my years buying software. I attempted to explain to her that this was nothing short of highway robbery and they have many competitors selling RADIUS software at 1/2 the price with yearly maintenance costs at a fraction of what she was saying. So my question to anyone else using 3Com's RADIUS software; is this true ? I am going to call Source Technology tomorrow to discuss this. She did mention that many otehrs have called her with the same confusion. Based upon what she is saying, I can see why. I also fail to see why someone would pay $1800+ for full hardware maintenance and phone support and then pay another $2K for new software. So am I out in left field on this ? Jeff Binkley ASA Network Computing
Subject: Re: (usr-tc) Open code
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-09 21:13:14
At 02:56 PM 2/9/98 -0700, Pete Ashdown wrote: >As I would presume the hardware companies wouldn't put the schematics out >for their hardware. Once again, all the access-server companies make their >money off of the hardware. The software is incidental. > The software is nearly *everything*!! Precisely why it should be in everyone's hands.. ;) I am fairly confident that any NAS provider has more money/time rapped up in software than harware design (including modem code) Precisely why they need the entire planet to help them. (for free) If they won't release their code, then give us memory maps, API's, developer services, or anything that would promote third-party software developement. Why not? Works for Apple, IBM, Microsoft, and just about all OS providers.. Might make em some money... Software sells hardware.. Even modems nowadays are software UART's and mod/demodulators that used to be done with hardware and are now DSP and microchip code. Now if *this* code was GNU based, we would have no modem wars.. no hype.. only lots of multi-compatible modems... BTW, I suspect there is more software technology in a new sportster than is in any terminal server or router.. Modems are a marvel of technical perversion. Alot of coolness goes into making something do something it was never meant to.. Nothing quite like it.. thanks to goverments/utilities inability to change old systems that make lots of money, usr gets rich with x2 also!! Good job! allen P.S. Interesting that my first computer (appleII '78) came complete with schematic.. I borrowed the jfet speaker circuit to put on an 8080 kit (only leds and 0-F keypad!) and "hand assembled" a program to play Dixie. (my prof was a rebel!! got an "A") At the time, apple grew like wildfire. Wunder what their problem is now?? The "red book" even listed woz's asm for the floppy and serial circuits.. But now they are stuggling. Could they have been too proprietary for their own (longterm) good??.. Is NAS (and the internet) in it's 'model-t' days too like they were? Apple's aggressive partnership with developers (early mac days) and the GNU project are the only two models I've ever seen successfully fill up the software "basket" (and sell computers) since the early microsoft domination.. But microsoft usually wins which explains things now doesn't it.. Sorry for rambling.. Like to tell the "hand assembled" story ;) am _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Netserver Config Example
From: Brian <signal@shreve.net>
Date: 1998-02-09 21:20:33
On Mon, 9 Feb 1998, Jeremy Lingmann wrote: > Brian, > > Some questions on the following commands: > > > set enh_routing on # used enhanced routing > > add ippool dialup1 208.214.44.1 62 # to aggregate routes > > Are you saying that by enabling both enh_routing and defining an ippool > the Total Control will then aggregate it's RIP route announcements? Yes, this was brought to the attention of the list by Michael Mittlestadt (sp?). You turn on enh_routing, define an ippool, and make a netmasks table entry for this to happen, and it works nicely, one route, per netserver, rather than 50+ routes. > > I've been trying to find a way to do this for quite some time. However, > from the documentation in the release notes on these two commands, I don't > see anything pertaining to that... Any comments? > Well, 3com didn't really explain it well, or explain it at all, which doesnt make alot of sense, since it no doubt makes there equiptment more attractive, even if it is rip, and it would make alot of people happy to be doing this. > (Just for clarification, I'm thinking of the word "aggregate" in terms of > the netserver announcing all of it's assigned IP addresses through a > single *ie* /26 RIP announcement, instead of a million individual /32 > announcements.) > Yes thats it :), if you have a /26 as your dialup pool, then the netserver will announce a /26 rather than 64 /32's :) I am glad the posting of my config helped at least one person see something I'm doing and they aren't, now I hope someone is going to show me something I could be doing better :) > -jeremy > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Call Me Stupid
From: Brian <signal@shreve.net>
Date: 1998-02-09 21:26:13
> > So my question to anyone else using 3Com's RADIUS software; is this true ? I > am going to call Source Technology tomorrow to discuss this. She did mention > that many otehrs have called her with the same confusion. Based upon what she > is saying, I can see why. I also fail to see why someone would pay $1800+ for > full hardware maintenance and phone support and then pay another $2K for new > software. > > So am I out in left field on this ? > > Jeff Binkley > ASA Network Computing If its like she says it is, that really sucks. I can't imagine a company screwing its customers over like that. I would think you buy the software, and "support contract" which would give you access to total service for one year, and get you updated software. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) MPIP & ISDN call termination
From: Brian <signal@shreve.net>
Date: 1998-02-09 21:29:35
On Mon, 9 Feb 1998, Jeremy Lingmann wrote: > I have two fairly quick (and technical) questions concerning the Total > Control system.... > > First of all, does Netserver MPIP (multi-chassis multilink ppp) work when > ISDN calls are terminated on the latest version of the Quad I-Modem cards? yes. > We are currently terminating all of our ISDN calls on Netserver cards (7 > chassis connected to a total of 14 PRI's) and for performance/monitoring > reasons we'd like to change it so that the Modem cards take over the load. > If this isn't supported, is it being worked on at all? thats one of the reasons we moved to the modems and away from the Munich. This way you can write one SNMP query which will work for isdn and analog calls. > > Thanks, > -jeremy > -- > .------------------------------------------------------------------. > | Jeremy Lingmann \ http://www | jeremy@europa.com | > | Network Manager \ .europa | fax: 503-796-9134 | > | Europa Communications, Inc. \ .com | phone: 503-222-9508 | > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Call Me Stupid
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-09 22:32:29
On Mon, 9 Feb 1998, Jeff Binkley wrote: > So my question to anyone else using 3Com's RADIUS software; is this true ? I > am going to call Source Technology tomorrow to discuss this. She did mention > that many otehrs have called her with the same confusion. Based upon what she > is saying, I can see why. I also fail to see why someone would pay $1800+ for > full hardware maintenance and phone support and then pay another $2K for new > software. > > So am I out in left field on this ? Heh. No. $2K is outrageous for a piece of simple server software. There are numerous *free* solutions available, complete with source. Not sure about NT support, but .. in some of our eyes, that's a Good Thing<tm>. <big grin> Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Netserver 8/I, radius, MPPP (fwd)
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-09 23:25:16
On Tue, 10 Feb 1998, Netlink Support Personnel wrote: > I didn't get any response from this. > I was unable to find anything in the manual that came with the netserver. > > Can someone at least point me to a web site or a document/manual that I > may not have on how to do this? > > Thank You > > ---------- Forwarded message ---------- > Date: Wed, 4 Feb 1998 10:43:03 -0600 (CST) > From: Netlink Support Personnel <support@ns1.netlinkcom.com> > To: usr-tc@lists.xmission.com > Subject: Netserver 8/I, radius, MPPP > > What do I need to do to set up Multilink PPP on a Netserver 8/I? > > I am authenticating with radius and can set the ISDN port limit in radius > to limit the user to 2 ports, but I am unsure of how to bond the 2 ports > together with IP address. > There is nothing special to do with MPpp with the NETServer 8/16. In 3.x release code it is automatic, you have a switch called set mp_adv on to make sure that you enable the same. In the 4.x code what you need to do is look at the setup of default user - the default user is the user parameters that is choosed if radius does not mention those parameters. If you do a show user default you will see the max-b channels for the user. If it is set to 1 then you will not get multi-link ppp. That is all to it. krish > Thanks, > Curt > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) Open code
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-09 23:38:02
At 02:33 PM 2/9/98 -0800, MegaZone wrote: >THIS IS A >COMPETITIVE INDUSTRY. Precisely.. And every person, company, and technology that has lasted over the years in the microcomputer industry (not nas) did so by sharing the burden of software developement on whatever hardware platform.. proprietary or open hardware, doesn't matter. If you want to make money for a long time, then have 100+ times more software than you alone could possibly develop inhouse. ComOS is an OS or sorts I'm assuming. Why wouldn't you want 1000 genius's developing for it... Offer developer services, info, even hardware discounts (for r&d only).. Would you sell less or more? I think more.. works for mainframes.. works for workstations.. works for super-nentendo.. Why not nas? >Do you understand that Grasshopper? Can you for >a moment grasp that these very differences are what feeds thousands of >people and keeps them employeed? "keeps them employeed?" Are they that lame??. scared?? greedy?? lazy?? or just egocentric... >That the competition is what drives >the creative process for most of them to create better ways of doing things - >so that they can sell more? No. there would be no competion in this industry if it wasn't for the "thevery" of Steve Jobs, Bill Gates and others.. Because we had some "openness" in the past, there is standardization and competition today. All our customers use open or cloned systems.. and get more bang for the buck than we do with nas.. As well as pay our bills (and yours).. Thank God for the sharing that has occured in the computer industry.. Might not be good for every single person, but it's good for the consumer.. >If not, you aren't operating in reality. I disagree. I am sure you have more knowledge or experience with NAS, than many do (and me), but you are limited there. Have you every personally developed any software yourself? marketed it? sold it to nasa, harvard, dartmouth, ucla, and even the king of kings.. disneyworld?? heh All over the world.. Lets talk magazine reviews and awards.. I doubt it. So cut out the grasshopper shit. I like to be cute too but for phun not for *ego*. >>Real proof is the key. Why would anyone launch a lawsuit over a suspicion? > >Thanks for proving my point. Vendor X borrows a concept from Vendor Y, >but Vendor Y can't prove it. Thuise Vendor Y is screwed. You forget that $$ was not the driving force behind GNU. Apply free enterprise to the princple, And I believe it strengthens it. Then with $$, you can afford to sue.. >>Besides, whoever gives out the source first would have the advantage. >>While other vendors are playing catch-up in copying the latest version, >>even better features and optimizations are being added. yeah that would suck for us MZ... See, whats good fer *you* ain't good fer me.. The thing is there are more consumers of nas than folks working at nas vendors. >If one feature were all there were - *maybe*. But what if Vendor X has >super feature Fu, and Vendor Y has super feature Bar. Vendor Y gives out >their code, and Vendor X improves their version of Bar to match, or maybe >surpass, Vendor Y. Now X has A better Fu, and an equal or better Bar - >while Vendor Y is still fighting to do the R&D needed to improve their Fu. >The competitive advantage is definitely with X here. Yeah if X has horns all over him... Hopefully consumers will not give money to those who don't reciprocate to the planet.. I am assuming an educated consumer here and some are not. So Packard Bell stays in business another year or two... I'm talking longterm here. Not some code release lasting a month.. >And being second with a feature is not bad. It has been Livingston's >philosophy on many things. Let someone else walk into the minefield first. >We have the engineering talent to be able to rapidly develop features and >deploy them in a solid form. So watch what X does, go "Ooh, THAT was a >mistake" and then avoid it. Many times you can come to market with a >feature a little later, but have it stable *before* the other guy because >they need to redo things. And redoing takes time. Very True. That's part of competiton. We don't live under socialism where we *have* to cooperate. We compete and it's "dog eat dog".. But even we "free people" have patents expire. Why? Sharing knowledge creates industries such as ours and even more competition.. And I'm not just talking 20 years here. If the nas vendors didn't "marry into the family", you guys might be in for longterm troubles.. Bet apple wished xerox or sun had bought them a year or two ago... >>for their hardware. Once again, all the access-server companies make their >>money off of the hardware. The software is incidental. > >And I already said this statement is completely false. As has already >been kindly pointed out. While I did say that the HW and SW for a system, >it is true that the software is modular. Most vendors use C for things >and write modular functions - not a new concept. We've stated openly >that we feel the major strength of the PortMaster product is ComOS - NOT >THE HARDWARE. Sure, the HW is good HW, but it isn't anything earth- >shaking in design. Few NAS products are. But what let's use use the >inexpensive HW and still get killer performance? *COMOS!* Without the >extremely tight coding in ComOS a PM would be lame. We make our money >from the software - pure and simple. One of the major reasons Lucent >targeted Livingston was ComOS. USR licensed ComOS too, recall. "past-tense"... apparently it didn't work out.. Now if the source were co-developed my all vendors and guru's alike, you guys would have to develop even better hardware to compete and have distinction in the marketplace.. We the consumer would have better hardware and software.. this looks bad on the short-term for nas vendors but is actually good in the long run as consumers are everything. Please us and you will make money. Don't and simply disappear (or get bought by some scared CEO with enough sense to be scared).. >Saying the software is incidental shows a complete lack of understanding >of the situation. > >>If you think the complexity of Navigator 4 compares to the complexity of an >>authentication server, you really don't have a grip on coding. Besides, The total developement for navigator stomps a NAS!.. It has too. Let's see... It's got apple's quicktime.. someones .jpg decoder.. audio, etc.. I bet apple has more in quicktime than any NAS vendor.. Netscape providing plugin api's helped them share the burden of keeping up until now.. Wish you guys thought the same... Ask yourself if you have a grip on coding.. (and long term survival) >You missed the point - not surprising after the others you missed. My >point is FOR NETSCAPE Navigator is like RADIUS is for Livingston. And >for Netscape, their Server products is like ComOS is for Livignston. > >Navigator and RADIUS are *enablers* - they help to sell the other products. yeah without them, there would be no market.. Creating a market then supplying it is pretty smart. This is done with openess not the other way.. >They add to the feature set, and help to spread the name and reputation of >the vendor. But they're not money makers in and of themselves. What they >do is help enable the sale of products tht *do* make money. For Netscape >that means helping to sell their server products, for Livignston that means >helping to sell PortMasters. web browsers have come farther in 18 months than NAS it would appear to me. Good for consumers.. good for you too. Lest, none of us have a job or market. >And that is the same reason for the License on 2.0 and up. Other vendors >were using our RADIUS to sell their products. "Well, we don't have a >RADIUS server, but you can use Livingston's free." So we stopped allowing >that. If you want a free server personally I recommend Cistron RADIUS, but >there are plenty to pick from. And this was smart? Ok MZ whatever you say.. I currently use *no* livingston product and contribute nil to your paycheck. How will you change that? By protecting yourselves? Don't think that'd do it. You gain customers by making them happy, and showing what you can do.. And if you've never used a disassembler in your whole life (and why would you need to use one) then why would you put down someones level of competency.. I bet you've "tried before you buyed" at least once in your life.. We are all potential customers, not just potential thieves.. >>we're not talking about the free redistribution of software. We're talking >>about giving customers access to the source. The whole reason I use > >I'll state it clearly: Customers cannot be trusted not to share the source >with another vendor. Precisely.. But Livingstone/lucent can be?? is that it?? >If you think that is extreme, we've already found customers redistributing >RADIUS source openly even though the license forbids it. And it is one >of the running jokes in the industry that if you tell someone something >under NDA it jsut takes a little longer for them to tell someone else. I've >had too many people send me things with "I shouldn't be telling you this, >but..." If you give the source to anyone, you increase your risk facotor. >If you give it to multiple people it is geometric. So just post it for everyone.. Then the cat is out of the bag, and progress can go forth unhendered.. It might hurt this or that months sales alittle, but I even doubt that much! >Just look at the problems Livingston/Lucent and USR/3Com have gone through >over the ComOS license. And that is a license between two major vendors! >You think anyone is eager to see it repeated with individuals on a mass >basis. yes.. Imagine it.. One nice OS and not five little puppies.. Bet we would have many of the features we wait months for. If you were bill gates advistor, we all couldn't use our sucky microsoft mice with x-windows... It would be the gravest violation. And I might use a logitech driver.. heavens to mergatriod.. >If you want full access to the source - Linux, GateD, and interface cards. >You can, and people do, roll their own access server. You are free to do >so, and no one will stop you. You can write your own source to your heart's >content and give it away to any and all who ask. You don't HAVE to buy an >access server from anyone. yeah but I like waiting.. Quake lag waiting.. Waiting for OSPF.. Even waiting for MPIP again.. Linux doesn't protect *real* jobs, so naturally I won't use it. Hate to think of hungry programmers.. Hit's home too much.. >We're perfectly willing to let the diehards go and do that - and I suspect >all of the major vendors are - because they make up less than 1/10th of 1% >of the marketplace. The majority of people don't have the talent, the time, >or the desire to build their own solutions. So they look to a vendor to >do the work for them. Don't count on that... Vendor's really suck at writing code.. I found that out when I bought a Lisa.. CSC Phd's suck too.. for probably the same reasons.. >I *could* build my own house, I used to do construction and know how to >do framing, wiring, plumbing, etc. But I'd probably hire a builder to do >it for me. I was screwed by my builder! So I will build or at least contract my next home.. You could not have picked a better example, for me anyway.. >I'm about to move cross country. I *could* pack a van myself, drive it >across country and unload it, but I'm having movers do it. Why? Because >I've DONE it myself and it is a big pain in the ass. I'd rather let >someone else carry an oak bedroom set around, deal with the drive, and >unload it. I saw this on catv once. That Given's guy sez to be successful, have others mow your lawn and stuff.. Cause you are worth more per hour. True but I cut lawns as a kid and like the smell of grass. (and I do much better job accordingly).. >I could go hunting for my meals - but I prefer to use the Safeway up the >street. I hunt/fish for pleasure too.. (before I became an ISP :) The resulting meal is much better than McDonalds.. but no cheaper... >You pay for the convenience and for expertise that you yourself might not >possess. If you can writing everything needed for an access server, from >Bellcore signalling to PPP to Stac to BGP - then you really should start >your own company. I too made the mistake of thinking I knew more than my customers.. And they needed me so bad I didn't have to release source code. I made alot of good customers mad only to release the source to public domain a couple of years later due to a lack thereof.. Only with apps does this work. extending a language for programmers without providing the source is a mistake IMHO.. It all depends on who your market it.. >>Merit's RADIUS server rather than Livingstons, or USRs, or Ascends is >>because *I* have control over the source, and *I* have made modifications >>(which anyone is free to have) that make it work better for *me*. > >And? You could do the same thing to Livingston's RADIUS - you're free to >redistribute diffs at the least. And Merit can always yank the rug out >by making major code changes between releases, so you can either split off >your tree, or rehack your features. > >>So far this discussion has generated three customers who exclaim how great >>the idea is, and one vendor who pooh-poohs it. right, it's good for consumers and bad for proprietary vendors in short term projections.. But it's good for both in the long haul I believe.. >What do you do for a living? Run an ISP? Why don't you give away your >dialin accounts. We do. They are called comps. Not for everyone. But for a significant percentage.. >Go on, let everyone in for free. Why not give your >user information out on the net, let all of your competitors have access. we have public guest accounts.. yes we have a 10 min limit. But at least you can see how we do things.. >What, doesn't sound like a good idea? I'd love a free, unlimited use >account with all the hardrive I can use. > >-MZ Yes our new local radio station does indeed like it. Cause it makes us both more money! allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Open code
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-10 00:17:15
At 04:13 PM 2/9/98 -0700, Pete Ashdown wrote: >MegaZone said once upon a time: > >>Come done from the clouds, or share what you're smoking. THIS IS A >>COMPETITIVE INDUSTRY. Do you understand that Grasshopper? Can you for >>a moment grasp that these very differences are what feeds thousands of >>people and keeps them employeed? > >Can you, for a single moment understand that customers are pissed >throughout this competitive industry? Bravo pete!!! My first post to this list implied how scarry it was and how much of a shock it was, to learn of the state of this industry. In 20 years dealing with hw/sw vendors (and being one as well), I have never seen such gross neglect of good customers. I can't believe that you (larger) guys sit still for it (except for a few remarks on this list..) Not that I am all that rebellious, just that hours on hold for support for a $20 accounting package is one thing, for a $36K purchase with support contract is another. (I know things have improved.. I'm talking way back when..) But it's more than that. It's UDP lag. It's ignoring your customer.. Call me old fashioned, but just because this stuff hooks to the telco, doesn't mean you have to act just like one.. We just dont like being at your total mercy anymore than the telco's.. (but they like it - hence their position, and MZ's) However, there is hope. The same competitive nature that presents the problem, will also in time, provide the solution. It always does.. allen
Subject: Re: (usr-tc) Open code
From: WCLynx SysOp <malevil@wclynx.com>
Date: 1998-02-10 04:21:37
Whew!~ This topic definitly hit a nerve... I'll try to summarize what the meat of the topic is. Let's see... Customers wish to see open source code to advance design and functionality of Vendor software. Vendors' view open source code as a threat to technical advantages with competitors in the product game. MZ, sorry, your point about source and schematic distribution only applies to recent developments in the industry. A kind a rampant a paranoia sweeping through. Microsoft probably has the greatest contributing factor in this problem. (My own father's company was doing development on OS2 up until last year. He has a copy of the entire source to OS2, w/ comments provided by IBM.) I personally have old zenith 386 tech manuals that have complete schematics and logic flow diagrams for both desktop and laptop computers. Other examples can and already have been cited. My personal opinion is the closed code influence came out of MS with their treatment of licensing. And it's a real pain. Refer to timing of the client licensing policy of MS Win NT and the policy change of developers like Software.com on their server products. No real change in the code, just a modified software key to allow more user connections. Sounds like Borg Greed to me. On the other side - Problems of the open code model... Bad code being released from developers who truly intend to help fix problem and add functionality. Without the checks and testing of the vendor, bugs become too prolific. CQ is a risk that the vendor will be held accountable for, regardless whether they created the problem code or not. This accountability affects their reputation and marketability. And finally the bottom line. I use NT here as our net OS and my parent company ap.net uses SUN Unix. I have been following the list carefully to see the USR-TC upgrade solutions using the Netservers, etc, for what's becoming a realm style management on our network and it's frightening. I've been using Digi boxes backing A/D quad modems for the past two years, and although I've seen improvements, the Digi's are the pinpoint weak link. But the problems I've had are flat out nothing compared to what I've been reading on this list. Now my boss at ap.net acquired another Portmaster 2e from one of his other sites for use here. OOoo yea... Now I get to integrate a new design into an already functional system. NT RAS on a 3.51 system has served admirably with few problems. Now I get to implement RADIUS. From all I've been reading here, it needs help, Major design and implementation help for communicating and serving all of the various hardware and OS models. And since my servers use the NT SAM for directory security, I need to learn how to marry the two database functions otherwise I get to do double user entries (one to Radius and one to NT) forever. And for the size of this installation, a move to Unix OS is just plain silly. NT is here, and we're finding more and more of our business customers moving to it. When we go to set-up the NT Dial-Up RAS, we hit hurdles. Our current Livingston RADIUS doesn't know what to make of the Domain login field, and dumps the connection. Hmm... Unless we've got old Radius code this looks like an integration problem with MS wins methodology to me. (BTW - the quick fix is to have the client clear the Domain name field in the Connect To dialog box.) As a developer, I'd like to help address and fix problems like these. Would open code solve my problems? Maybe, maybe not. But at least we would be moving faster towards better inter-system communications than we have now. As ISPs our business is Service. Anything we can do toward that end makes everyone's life a little easier. The last thing we need is petty infighting from our vendors. It seems they've lost sight of the end goal, that happy clients make for happy profits and growth in the industry overall. I do maintain that the vendor must control the final release of changes to hardware and the code to ensure quality and stability. This site gains clients with no advertising other than word of mouth because our ISP provides decent, knowledgeable, and sometimes just plain old moral support for our users. I'm ready for more modems, but the lack of true customer support is appalling. (Refer to previous messages on software upgrade restrictions, and usage terms.) Maybe we should go with String Net featuring the nonproprietary Cup and Square Knot connections for reliability, but a decent modem system would be preferred. I like the TC design but it's pricing itself out of our market on growth support alone. I've got 4 courier v.everythings in here because I couldn't justify the cost of just 1 quad A/D modem card set for the TC Hub. And the firmware code is more current than for the TC cards that I can't get the update for because "I don't qualify." X2 here, forget it. Even if the local phone lines could support it, just to get the code for the price heavy hardware I already have is rediculous. Am I missing something, or does 3com not want to be in the modem/terminal server business anymore? (rant, rave, fume) I heard an old saying once, "If you're not part of the solution, the you're part of the problem." We as an industry need to drive the solutions. BTW - I've got the source now from Livingston for the NT Radius server. Any assistance in implementing it with NT 4.0 OS SAM file security, the Portmaster 2e and TC hub A/D quad modems would be well appreciated. Dave Sherry Site Network Engineer AccessPort's West Coast Lynx 10940 River Road / Forestville, CA 95436 707-887-4640 (voice) 707-887-1810 (Fax) http://www.wclynx.com e-mail: Malevil@wclynx.com, hostmaster@ap.net, support@wclynx.com, Webmaster@wclynx.com You've been Geeking to long if, you think in tag lines.
Subject: (usr-tc) Netserver 8/I, radius, MPPP (fwd)
From: Netlink Support Personnel <support@ns1.netlinkcom.com>
Date: 1998-02-10 07:51:06
I didn't get any response from this. I was unable to find anything in the manual that came with the netserver. Can someone at least point me to a web site or a document/manual that I may not have on how to do this? Thank You ---------- Forwarded message ---------- What do I need to do to set up Multilink PPP on a Netserver 8/I? I am authenticating with radius and can set the ISDN port limit in radius to limit the user to 2 ports, but I am unsure of how to bond the 2 ports together with IP address. Thanks, Curt
Subject: RE: (usr-tc) Open code
From: Tom Bilan <tom@tdi.net>
Date: 1998-02-10 07:53:46
Check out RadiusNT from iea-software.com, I believe it works with the NT SAM database. I moved from Linux to NT and although I have modified several programs, like Radius from Livingston, to fit my own purposes I still moved to NT. Why? For $195 I can make any problem Microsoft's problem and they WILL fix it. I had an exchange server problem about 3 weeks ago and although I find myself on the top of the food chain when it comes to fixing things, M.S. CSE and MCSE, I couldn't pinpoint the problem. I went through 4 techs at Micro$oft before they finally woke some tech up at home, who I now affectionally call Exchange Superman, and he had me up and running in < 1 hour. Try that with the open source of Linux. Not that I'm not on the side of open code... Tom Bilan TDI Internet > -----Original Message----- > From: WCLynx SysOp [SMTP:Malevil@wclynx.com] > Sent: Tuesday, February 10, 1998 6:22 AM > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) Open code > > Whew!~ > > This topic definitly hit a nerve... I'll try to summarize what > the meat of the topic is. Let's see... > > Customers wish to see open source code to advance design and > functionality of > Vendor software. > > Vendors' view open source code as a threat to technical advantages > with > competitors in the product game. > > MZ, sorry, your point about source and schematic distribution only > applies to > recent developments in the industry. A kind a rampant a paranoia > sweeping > through. Microsoft probably has the greatest contributing factor in > this > problem. (My own father's company was doing development on OS2 up > until last > year. He has a copy of the entire source to OS2, w/ comments provided > by IBM.) > I personally have old zenith 386 tech manuals that have complete > schematics and > logic flow diagrams for both desktop and laptop computers. Other > examples can > and already have been cited. My personal opinion is the closed code > influence > came out of MS with their treatment of licensing. And it's a real > pain. Refer > to timing of the client licensing policy of MS Win NT and the policy > change of > developers like Software.com on their server products. No real change > in the > code, just a modified software key to allow more user connections. > Sounds like > Borg Greed to me. > > On the other side - Problems of the open code model... Bad code being > released from developers who truly intend to help fix problem and add > functionality. Without the checks and testing of the vendor, bugs > become too > prolific. CQ is a risk that the vendor will be held accountable for, > regardless whether they created the problem code or not. This > accountability > affects their reputation and marketability. And finally the bottom > line. > > I use NT here as our net OS and my parent company ap.net uses SUN > Unix. I > have been following the list carefully to see the USR-TC upgrade > solutions > using the Netservers, etc, for what's becoming a realm style > management on our > network and it's frightening. > > I've been using Digi boxes backing A/D quad modems for the past two > years, and > although I've seen improvements, the Digi's are the pinpoint weak > link. But the > problems I've had are flat out nothing compared to what I've been > reading on > this list. Now my boss at ap.net acquired another Portmaster 2e from > one of > his other sites for use here. OOoo yea... Now I get to integrate a > new design > into an already functional system. NT RAS on a 3.51 system has served > admirably with few problems. Now I get to implement RADIUS. From all > I've > been reading here, it needs help, Major design and implementation help > for > communicating and serving all of the various hardware and OS models. > And since > my servers use the NT SAM for directory security, I need to learn how > to marry > the two database functions otherwise I get to do double user entries > (one to > Radius and one to NT) forever. And for the size of this installation, > a move > to Unix OS is just plain silly. > > NT is here, and we're finding more and more of our business customers > moving > to it. When we go to set-up the NT Dial-Up RAS, we hit hurdles. Our > current > Livingston RADIUS doesn't know what to make of the Domain login field, > and > dumps the connection. Hmm... Unless we've got old Radius code this > looks like > an integration problem with MS wins methodology to me. (BTW - the > quick fix is > to have the client clear the Domain name field in the Connect To > dialog box.) > As a developer, I'd like to help address and fix problems like these. > > Would open code solve my problems? Maybe, maybe not. But at least we > would be > moving faster towards better inter-system communications than we have > now. As > ISPs our business is Service. Anything we can do toward that end > makes > everyone's life a little easier. The last thing we need is petty > infighting > from our vendors. It seems they've lost sight of the end goal, that > happy > clients make for happy profits and growth in the industry overall. I > do > maintain that the vendor must control the final release of changes to > hardware > and the code to ensure quality and stability. This site gains clients > with no > advertising other than word of mouth because our ISP provides decent, > knowledgeable, and sometimes just plain old moral support for our > users. I'm > ready for more modems, but the lack of true customer support is > appalling. > (Refer to previous messages on software upgrade restrictions, and > usage terms.) > > Maybe we should go with String Net featuring the nonproprietary Cup > and Square > Knot connections for reliability, but a decent modem system would be > preferred. > I like the TC design but it's pricing itself out of our market on > growth > support alone. I've got 4 courier v.everythings in here because I > couldn't > justify the cost of just 1 quad A/D modem card set for the TC Hub. And > the > firmware code is more current than for the TC cards that I can't get > the update > for because "I don't qualify." X2 here, forget it. Even if the local > phone > lines could support it, just to get the code for the price heavy > hardware I > already have is rediculous. Am I missing something, or does 3com not > want to > be in the modem/terminal server business anymore? (rant, rave, fume) > I heard an old saying once, "If you're not part of the solution, the > you're > part of the problem." We as an industry need to drive the solutions. > > BTW - I've got the source now from Livingston for the NT Radius > server. Any > assistance in implementing it with NT 4.0 OS SAM file security, the > Portmaster > 2e and TC hub A/D quad modems would be well appreciated. > > Dave Sherry > Site Network Engineer > > AccessPort's West Coast Lynx > 10940 River Road / Forestville, CA 95436 > 707-887-4640 (voice) 707-887-1810 (Fax) > > http://www.wclynx.com > e-mail: Malevil@wclynx.com, hostmaster@ap.net, > support@wclynx.com, Webmaster@wclynx.com > > You've been Geeking to long if, > you think in tag lines. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) Problems with HiperDsp
From: Brian <signal@shreve.net>
Date: 1998-02-10 07:54:07
On Tue, 10 Feb 1998, Richard Roy wrote: > We are using HiperDSP card with v.1.0.7 in a non-PRI mode. The problems are: > - The card forget it's modem + dsp setting and goes back to it's default > value after a reboot > - Sometimes the DSP disapear and we have to reboot to see it back. > > I'm wandering who are using the HiperDSP card and if those problems are > already fixed. I am assuming you did a cmd svpcfg > > -------------------------------------------------------------------------- > Richard Roy (rjroy@nbnet.nb.ca) > NBTel Internet > Technical Analyst / Analyste Technique > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Problems with HiperDsp
From: Richard Roy <rjroy@nbnet.nb.ca>
Date: 1998-02-10 09:30:23
We are using HiperDSP card with v.1.0.7 in a non-PRI mode. The problems are: - The card forget it's modem + dsp setting and goes back to it's default value after a reboot - Sometimes the DSP disapear and we have to reboot to see it back. I'm wandering who are using the HiperDSP card and if those problems are already fixed. Richard Roy (rjroy@nbnet.nb.ca) NBTel Internet Technical Analyst / Analyste Technique
Subject: Re: (usr-tc) blocking non x2 subscribers...
From: jason_kelton@3com.com
Date: 1998-02-10 09:57:23
My suggestion would be to create multiple DNIS groups, disable X2 on one group (enabling all other protocols), enable X2 on the other group (disabling all other connection rates), then use a DNIS restrictions list on your auth server. Regards, Jason Kelton Carrier Systems Engineer 3Com ANZA. db3l@ans.net on 10/02/98 07:35:30 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) Ian Roy <iroy@smokies.amnix.com> writes: > I like your solutions above. But, I have one question: > > What happens to a user who has not subscribed to X2 services, but > has an X2 modem ? Assuming that the phrase "X2 services" refers to some sort of service model or access plan that a user must sign up for specifically, then obviously the modems can't tell - as long as the user has an x2 capable client modem they'll be allowed to train successfully by the server modems. I may have misinterpreted "X2 subscribers" in Matthew's original request to mean "subscribers who have an x2 modem" rather than users who subscribe to a separate x2-supporting service. In the former case - where you offer x2 services to your subscribers who have an appropriate modem but don't want non-x2 modems to tie up your better quality lines - controlling it at the modem should work fine. But in the latter case - where you need to know if the user is paying for x2 service - it's not sufficient and you will need to do something at some other layer such as authentication. I suppose the modem restrictions might be marginally helpful even in the separate service case in terms of throttling back on the number of requests you need to work on and/or helping to get rid of non-x2 people faster than authentication or a periodic scan might be able to. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/ - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) Problems with HiperDsp
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-10 10:10:40
: We are using HiperDSP card with v.1.0.7 in a non-PRI mode. The problems are: : - The card forget it's modem + dsp setting and goes back to it's default : value after a reboot : - Sometimes the DSP disapear and we have to reboot to see it back. The nice people at 3Com put 1.0.95 on our HDM, and it's run like a champ now for a week. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: Re: (usr-tc) Call Me Stupid
From: Beth Montes <bmontes@iland.net>
Date: 1998-02-10 10:42:25
For the most part, 3Com's tech support people will unlock one or two files for you so that you can get an update. They tend to be fair-minded and do what's best for the customer in most situations. This is not to say that they ignore their corporate policy (misguided though it may be) of having legitimate customers pay for every tiny modification in code, but they *do* evaluate each situation on its own merits and make intelligent decisions. In my experience, it's best to make your requests to a knowledgable tech support person, and not get shoved through the system to some paperwork type person who will read corporate policy to you. Beth _____________________________________________ Beth Montes bmontes@iland.net 816-827-5111 x21 ComputerLand of Sedalia http://www.c-land.com I-Land Internet Services http://www.iland.net _____________________________________________ At 08:31 PM 2/9/98 -0500, Jeff Binkley wrote: > >Ok folks. I thought I could understand just about anything but after a >conversation with a 3Com representative, my head is spinning and I am totally >confused. Ok, here's the situation. > >Last year we bought one of the ISP bundles from USR, 1706 I believe, which >included the Total Control Chassis, TCM software and the Windows based security >and accounting server software (i.e. RADIUS). After 90 days, I lost access to >all of the updates on the TotalService website (with no prior notice mind you) >and a quick call to now 3Com, I was told we needed a support agreement. Ok, >next quick call to Source Technology and I am told we need the PCAP agreement, >so we shelled out $1800+ and got the highest agreement they had. Up until now >I was under the impression we were all set. Now January comes along and the >new RADIUS software gets posted on the Total Service website. A call to the >support line sends me to someone else at 3Com who handles access to the site. >Now comes the confusion. > >In talking to a new lady there she attempts to explain to me that I don't have >what I need to get RADIUS updates. So I attempt to ask her why ? She says the >ISP bundles "run out" after a certain amount of time. I tried to get her to >explain what she means. She then goes on to say the RADIUS software isn't >included under the CAP agreement and what I really need is part number >001611-00, which ends up being the RADIUS software. Now, in my obvious >confusion, I ask her why I should buy it again, since it was included in the >bundle, which in my mind is nothing more than a sale price instead of me buying >the parts individually. She then again says that it doesn't work this way. So >I ask her if I buy the RADIUS software again for the nearly $2K list (according >to her, I haven't verified this price yet), what happens at the end of another >year ? Is there a software maintenance agreement I need to get. Again, she >says "No". At the end of the year I need to purchase the software again ??? >Now, call me stupid but I have never heard of such a thing in all of my years >buying software. I attempted to explain to her that this was nothing short of >highway robbery and they have many competitors selling RADIUS software at 1/2 >the price with yearly maintenance costs at a fraction of what she was saying. > >So my question to anyone else using 3Com's RADIUS software; is this true ? I >am going to call Source Technology tomorrow to discuss this. She did mention >that many otehrs have called her with the same confusion. Based upon what she >is saying, I can see why. I also fail to see why someone would pay $1800+ for >full hardware maintenance and phone support and then pay another $2K for new >software. > >So am I out in left field on this ? > >Jeff Binkley >ASA Network Computing > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > >
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-10 10:48:02
Check it out on www.3com.com krish On Tue, 10 Feb 1998, Lee Kuo wrote: > On Tue, 10 Feb 1998, Garry Shtern wrote: > > When are you guys releasing the ITU 56k code update? > > Hate to be lame, but "me too". I'm sure that everyone else who owns any > USR/3COM equipment would very much like to know the answer to this. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) Call Me Stupid
From: Brian <signal@shreve.net>
Date: 1998-02-10 10:51:57
On Tue, 10 Feb 1998, Beth Montes wrote: > For the most part, 3Com's tech support people will > unlock one or two files for you so that you can > get an update. They tend to be fair-minded and do > what's best for the customer in most situations. > This is not to say that they ignore their corporate > policy (misguided though it may be) of having > legitimate customers pay for every tiny modification > in code, but they *do* evaluate each situation on > its own merits and make intelligent decisions. > > In my experience, it's best to make your requests > to a knowledgable tech support person, and not get > shoved through the system to some paperwork type > person who will read corporate policy to you. > > Beth I aggree, here is a lame situation: Customer pays for Netserver v3.3.28, which claims to have support for "feature foobar" Netserver feature foobar is broken, and doesnt do what its suppose to do. Customers support runs out Netserver v3.4.x comes out, feature foo works in this version Customer is left with something he paid for, but does not work. Customer checks the USR inventory packing list, and finds that indeed no vasaline was included in the sale. > _____________________________________________ > > Beth Montes > bmontes@iland.net 816-827-5111 x21 > ComputerLand of Sedalia http://www.c-land.com > I-Land Internet Services http://www.iland.net > _____________________________________________ > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Problems with HiperDsp
From: Richard Roy <rjroy@nbnet.nb.ca>
Date: 1998-02-10 11:05:59
At 07:54 AM 2/10/98 -0600, you wrote: >On Tue, 10 Feb 1998, Richard Roy wrote: > >> We are using HiperDSP card with v.1.0.7 in a non-PRI mode. The problems are: >> - The card forget it's modem + dsp setting and goes back to it's default >> value after a reboot >> - Sometimes the DSP disapear and we have to reboot to see it back. >> >> I'm wandering who are using the HiperDSP card and if those problems are >> already fixed. > > >I am assuming you did a cmd svpcfg Yes. USR told us that they were aware of this problem. We use HiperArc 4.0.80 I presume you dont have this kind of problem. Richard Roy (rjroy@nbnet.nb.ca) NBTel Internet Technical Analyst / Analyste Technique
Subject: (usr-tc) New Netserver code out yesterday
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-10 11:33:00
The new Netserver code came out on the 9th. Version 3.7.24.
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Brian <signal@shreve.net>
Date: 1998-02-10 13:28:06
On Tue, 10 Feb 1998, Pete Ashdown wrote: > The new Netserver code came out on the 9th. Version 3.7.24. > new netserver code new quad modem code so whos installing first? whos the guinea pig :)) Brian > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-10 13:38:16
> new netserver code > new quad modem code > > so whos installing first? whos the guinea pig :)) Very little is changed between 3.6.28 and 3.7.24. It is *supposed* to help with the Quake lag, and some other minor bugs, but I've been running 3.7.24 for weeks now with no problems at all. I'm also now busy flashing all my chassis with the latest modem code. During the beta run, there were no problems on my beta box, however. I've been running it for almost a week on production boxes with no problems. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Netserver Config Example
From: Mike Wronski <mwronski@coredump.ae.usr.com>
Date: 1998-02-10 13:56:29
On Mon, 9 Feb 1998, Brian wrote: .... > add mpipclient 208.206.76.41 secret > save all # Save to NVRAM <wait for save to finish> > reboot # reboot -m
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Charles Hill <chill@ionet.net>
Date: 1998-02-10 13:57:14
On Tue, 10 Feb 1998, Brian wrote: > On Tue, 10 Feb 1998, Pete Ashdown wrote: > > > The new Netserver code came out on the 9th. Version 3.7.24. > > new netserver code > new quad modem code > > so whos installing first? whos the guinea pig :)) > Brian We've been running 3.7.24 since last Thursday on 3 Netservers and have begun deploying it system-wide now that the coast is clear. It's good stuff. The redesigned MPIP is supposed to be much more reliable during call setup, but since this MPIP doesn't interoperate with the TCS 2.5 MPIP we will have to upgrade Netservers at each POP all at once. We haven't finished testing the quad code. -CH
Subject: Re: (usr-tc) Netserver Config Example
From: Charles Hill <chill@ionet.net>
Date: 1998-02-10 14:27:09
> On Mon, 9 Feb 1998, Brian wrote: > .... > > add mpipclient 208.206.76.41 secret > > save all # Save to NVRAM > <wait for save to finish> > > reboot # reboot > -m Good call, Mike. I've been bitten by that one! I've done save all followed quickly by a reboot and had to busy out the PRI lines and reverse telnet to the console port to re-enter the settings. -CH
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Jaye Mathisen <mrcpu@cdsnet.net>
Date: 1998-02-10 14:35:53
If you read the release notes, you'll see some notes about quake UDP fixes. however, it does not specifically state they came out of 3.6.69 On Tue, 10 Feb 1998, Brian Elfert wrote: > > > On Tue, 10 Feb 1998, Pete Ashdown wrote: > > > The new Netserver code came out on the 9th. Version 3.7.24. > > Does anyone know if this includes the Quake fixes supposedly in 3.6.69? > > Brian > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Jaye Mathisen <mrcpu@cdsnet.net>
Date: 1998-02-10 14:35:53
If you read the release notes, you'll see some notes about quake UDP fixes. however, it does not specifically state they came out of 3.6.69 On Tue, 10 Feb 1998, Brian Elfert wrote: > > > On Tue, 10 Feb 1998, Pete Ashdown wrote: > > > The new Netserver code came out on the 9th. Version 3.7.24. > > Does anyone know if this includes the Quake fixes supposedly in 3.6.69? > > Brian > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) User cant get on
From: Brian <signal@shreve.net>
Date: 1998-02-10 14:51:20
We have users that cant get on our Hubs, I did a PPP dump on them and get this: Outgoing PPP Data on interface: slot:10/mod:21 PAP ACK Outgoing PPP Data on interface: slot:10/mod:21 IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 NEW_ADDRS d0 ce 4c 23 Incoming PPP Data on interface: slot:10/mod:21 CCP CFG_REQ MS_COMP 00 00 00 01 STAC_COMP 00 01 04 Outgoing PPP Data on interface: slot:10/mod:21 CCP CFG_REJ STAC_COMP 00 01 04 Outgoing PPP Data on interface: slot:10/mod:21 CCP CFG_REQ MS_COMP 00 00 00 01 Incoming PPP Data on interface: slot:10/mod:21 LCP PROT_REJ IPCP COMPR_TYPE 00 2d 0f 00 NEW_ADDRS d0 ce 4c 23 They are running Win95, latest DUN, compression enabled, log on to netowrk is off, and everything is server assigned....only about 4 users that cant get on. To me it looks like they are rejecting compression, and the new ip address..........any ideas? Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) User cant get on
From: chris loelke <chris@okanagan.net>
Date: 1998-02-10 15:01:45
had a similar problem with NT users and the latest dun turn MS compression off. At 02:51 PM 2/10/98 -0600, you wrote: >We have users that cant get on our Hubs, I did a PPP dump on them and get >this: > >Outgoing PPP Data on interface: slot:10/mod:21 > PAP ACK >Outgoing PPP Data on interface: slot:10/mod:21 > IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS d0 ce 4c 23 > >Incoming PPP Data on interface: slot:10/mod:21 > CCP CFG_REQ MS_COMP 00 00 00 01 > STAC_COMP 00 01 04 > >Outgoing PPP Data on interface: slot:10/mod:21 > CCP CFG_REJ STAC_COMP 00 01 04 > >Outgoing PPP Data on interface: slot:10/mod:21 > CCP CFG_REQ MS_COMP 00 00 00 01 > >Incoming PPP Data on interface: slot:10/mod:21 > LCP PROT_REJ IPCP > COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS d0 ce 4c 23 > > >They are running Win95, latest DUN, compression enabled, log on to netowrk >is off, and everything is server assigned....only about 4 users that cant >get on. > >To me it looks like they are rejecting compression, and the new ip >address..........any ideas? > >Brian > > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Network Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Brian <signal@shreve.net>
Date: 1998-02-10 15:17:20
On Tue, 10 Feb 1998, Curt Shambeau wrote: > > new netserver code > > new quad modem code > > > > so whos installing first? whos the guinea pig :)) > > Very little is changed between 3.6.28 and 3.7.24. It is *supposed* to > help with the Quake lag, and some other minor bugs, but I've been running > 3.7.24 for weeks now with no problems at all. I'm also now busy flashing The release notes was like 3 pages long. All fixes which is good, as long as nothing broke in the fixing of something. > all my chassis with the latest modem code. During the beta run, there > were no problems on my beta box, however. I've been running it for almost > a week on production boxes with no problems. Are you putting the quad code on chassis that do not have the new netserver code? Is it safe to put the new quad code on 3.6.x chassis? Brian > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-10 15:26:46
> > all my chassis with the latest modem code. During the beta run, there > > were no problems on my beta box, however. I've been running it for almost > > a week on production boxes with no problems. > > Are you putting the quad code on chassis that do not have the new > netserver code? Is it safe to put the new quad code on 3.6.x chassis? No, I'm pretty sure they do no recommend running the new Quad code on the older 3.6.28 netserver code. I'm running the new code on 3.7.24. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-10 16:26:54
On Tue, 10 Feb 1998, Pete Ashdown wrote: > The new Netserver code came out on the 9th. Version 3.7.24. Does anyone know if this includes the Quake fixes supposedly in 3.6.69? Brian
Subject: Re: (usr-tc) NMC card code
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-10 16:50:00
> I have an NMC card in a fairly new chassis that, when pulled up in TCM, > is asking for hm*.nac code to upgrade it. I can't find this code > anywhere (TCM seems to be label it as NMC code (8 MB) or something like > that. Any idea what this thing is asking for? Any NMC code that starts with HM is for the Hiper Chassis. It means the NMC has already been upgraded with RAM and a larger flash ROM. You can pull the code off of totalservice - just look for the NMC code for 16MB NMC cards. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) NMC card code
From: Brian <signal@shreve.net>
Date: 1998-02-10 16:57:36
On Tue, 10 Feb 1998, Curt Shambeau wrote: > > I have an NMC card in a fairly new chassis that, when pulled up in TCM, > > is asking for hm*.nac code to upgrade it. I can't find this code > > anywhere (TCM seems to be label it as NMC code (8 MB) or something like > > that. Any idea what this thing is asking for? > > Any NMC code that starts with HM is for the Hiper Chassis. It means the > NMC has already been upgraded with RAM and a larger flash ROM. > > You can pull the code off of totalservice - just look for the NMC code for > 16MB NMC cards. Its the nmc522 code which is for 16MB cards like Curt said, the nmc532 is for 8mb cards. you *can* run the nmc522 (16mb) code on a nmc in a normal netserver/quad chassis, as we do. Brian > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) MPIP & ISDN call termination
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-10 17:10:39
Thus spake Jeremy Lingmann >I have two fairly quick (and technical) questions concerning the Total >Control system.... >First of all, does Netserver MPIP (multi-chassis multilink ppp) work From what I've seen, I can truncate this question to this point and say "no." Call reference number 27457. This is why we moved away from Livingston equipment, because we had an open issue with them that was *never* dealt with. This call has been open since Dec. 31. Sad. And, yes, the time is sync'ed on the chassis'. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Mike Wronski <mwronski@coredump.ae.usr.com>
Date: 1998-02-10 17:15:51
On Tue, 10 Feb 1998, Brian Elfert wrote: > > > On Tue, 10 Feb 1998, Pete Ashdown wrote: > > > The new Netserver code came out on the 9th. Version 3.7.24. > > Does anyone know if this includes the Quake fixes supposedly in 3.6.69? > All fixes from 3.6.69 were pushed into the 3.7.24 code. -M
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-10 17:25:38
On Tue, 10 Feb 1998, Garry Shtern wrote: > When are you guys releasing the ITU 56k code update? Hate to be lame, but "me too". I'm sure that everyone else who owns any USR/3COM equipment would very much like to know the answer to this.
Subject: (usr-tc) NMC card code
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-10 17:45:31
I have an NMC card in a fairly new chassis that, when pulled up in TCM, is asking for hm*.nac code to upgrade it. I can't find this code anywhere (TCM seems to be label it as NMC code (8 MB) or something like that. Any idea what this thing is asking for? -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-10 18:02:47
On Tue, 10 Feb 1998, Curt Shambeau wrote: > > > all my chassis with the latest modem code. During the beta run, there > > > were no problems on my beta box, however. I've been running it for almost > > > a week on production boxes with no problems. > > > > Are you putting the quad code on chassis that do not have the new > > netserver code? Is it safe to put the new quad code on 3.6.x chassis? > > No, I'm pretty sure they do no recommend running the new Quad code on the > older 3.6.28 netserver code. I'm wondering why exactly this new modem code requires new Netserver code? The old code revs have never need new Netserver code. Brian
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-10 18:23:37
At 05:15 PM 2/10/98 -0600, Mike Wronski wrote: >On Tue, 10 Feb 1998, Brian Elfert wrote: > >> >> >> On Tue, 10 Feb 1998, Pete Ashdown wrote: >> >> > The new Netserver code came out on the 9th. Version 3.7.24. >> >> Does anyone know if this includes the Quake fixes supposedly in 3.6.69? >> > >All fixes from 3.6.69 were pushed into the 3.7.24 code. > When are you guys releasing the ITU 56k code update? Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) NMC card code
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-10 21:47:02
On Tue, 10 Feb 1998, Curt Shambeau wrote: > > I have an NMC card in a fairly new chassis that, when pulled up in TCM, > > is asking for hm*.nac code to upgrade it. I can't find this code > > anywhere (TCM seems to be label it as NMC code (8 MB) or something like > > that. Any idea what this thing is asking for? > > Any NMC code that starts with HM is for the Hiper Chassis. It means the > NMC has already been upgraded with RAM and a larger flash ROM. Note that 3COM has been shipping the NMC cards with 16MB ram and 8MB (Or is it 4MB?) flash for some time now. I was suprised to see that my 2059 chassises purchased in Sept and Dec already have the NMC upgrades. Brian
Subject: Re: (usr-tc) Netserver route problem
From: Russ Panula <rpanula@dacmail.net>
Date: 1998-02-11 01:33:29
On Mon, 9 Feb 1998 15:19:27 -0600 (CST), Brian <signal@shreve.net> wrote: >> >yet, whenever I start gated on a machine (running rip), I get this in the >> >routing tables: >> > >> >208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 >> >> Try adding >> >> Framed-Netmask = 255.255.255.255 > >Why would I delcare a Framed-Netmask on a Framed-Route account? Specifying a host netmask cleared up an erroneous route advertisement I was seeing with hosts authenticated by similar RADIUS entries and routing /28's. I'm guessing usr2ts1 is announcing a /24 route because the client in question is being assigned inet 208.214.45.1 mask 255.255.255.0. What netmask is the client assigned? What routes are successfully announced when the client connects?
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-11 09:34:42
> OK, that brings up the opposite question - is it safe to run 3.7.x with > the OLD (5.5.7/5.6.7) Quad firmware? Yes, that works fine - at least it did here. I had several boxes running 48 ports of Quad modems with 5.6.7 + 2 HDM's running netserver code 3.7.24 for a month or so. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: (usr-tc) ITU Standard
From: Terry Kennedy <terry@olypen.com>
Date: 1998-02-11 10:10:53
So when USR (3com) puts out the upgrade to the new standard (v.pcm?) will all the customers we have that have bought sporters or other X2 modems need to upgrade their modems? This could be real nightmare... Or will be so cool as to still work with X2 products? After reading the white papers at 3com.com it was still unclear to me. Any clarifications would help. Thanks... Terry Kennedy System Admin OlyPen, Inc.
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Greg Coffey <greg@coffey.com>
Date: 1998-02-11 10:31:47
According to Eric Benhamou, the head honcho at 3Com, "The V.pcm standard is so close to x2 that 3Com expects to ship V.pcm-compliant products within days after the standard is voted on. "We'll provide a free [downloadable] software upagrade to our x2 customers who want to run V.pcm. As it stands today, the content of standard has been resolved, and the confididence of having a formal V.pcm modem standard is extremely high." At 12:11 PM 2/11/98 -0500, you wrote: >At 10:48 AM 2/10/98 -0600, Tatai SV Krishnan wrote: >>Check it out on www.3com.com >> >>krish > >I did.. But it did not say anything about Total Controls, I think it only >refered to modems, so I wanted to make sure ;) > >Garry Shtern shterng@akula.com >Chief Network Administrator http://www.akula.com >Akula Communications Corp. tel. (212) 292-8892 > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > Thanks, Greg Coffey, CoffeyNet ** $20 local Casper USR x2 56k access ** 142 S. Center St. Wheatland, Pinedale, Lander, Lusk Casper, WY 82601 Douglas & Rawlins (307) 234-5443 http://www.coffey.com Open 8-6 M-F / 10-2 Saturday
Subject: (usr-tc) Hiper ISDN problems
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-11 10:39:50
We have several customers with this problem, they connect just fine with the Total Control and the HiPerDSP cards, but are unable to surf, I traced several with the monitoring commands of the HiPerARC, and found some weird things. The interface asks for quite a lot of RESET's. Here's an example, if anyone has seen something like it and got a solution, please let me know. The customers are using Win95 without DUN 1.2, with a config that worked often. Thanks for any info, Robert von Bismarck Petrel Communication S.A. Outgoing PPP Data on interface: slot:3/mod:1 CHAP SUCCESS 00 Outgoing PPP Data on interface: slot:3/mod:1 IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 NEW_ADDRS c2 33 c5 e4 Incoming PPP Data on interface: slot:3/mod:1 CCP CFG_REQ STAC_COMP 00 01 03 Outgoing PPP Data on interface: slot:3/mod:1 CCP CFG_ACK STAC_COMP 00 01 03 Outgoing PPP Data on interface: slot:3/mod:1 CCP CFG_REQ STAC_COMP 00 01 03 Incoming PPP Data on interface: slot:3/mod:1 CCP CFG_ACK STAC_COMP 00 01 03 Incoming PPP Data on interface: slot:3/mod:1 IPCP CFG_REQ COMPR_TYPE 00 2d 0f 01 NEW_ADDRS 00 00 00 00 81 00 00 00 00 82 00 00 00 00 83 00 00 00 00 84 00 00 00 00 Outgoing PPP Data on interface: slot:3/mod:1 IPCP CFG_REJ 82 00 00 00 00 83 00 00 00 00 84 00 00 00 00 Incoming PPP Data on interface: slot:3/mod:1 IPCP CFG_REQ COMPR_TYPE 00 2d 0f 01 NEW_ADDRS 00 00 00 00 81 00 00 00 00 Outgoing PPP Data on interface: slot:3/mod:1 IPCP CFG_NAK COMPR_TYPE 00 2d 0f 00 NEW_ADDRS 90 55 62 22 81 c2 33 c5 f0 Incoming PPP Data on interface: slot:3/mod:1 IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 NEW_ADDRS 90 55 62 22 81 c2 33 c5 f0 Outgoing PPP Data on interface: slot:3/mod:1 IPCP CFG_ACK COMPR_TYPE 00 2d 0f 00 NEW_ADDRS 90 55 62 22 81 c2 33 c5 f0 Outgoing PPP Data on interface: slot:3/mod:1 IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 NEW_ADDRS c2 33 c5 e4 Incoming PPP Data on interface: slot:3/mod:1 IPCP CFG_ACK COMPR_TYPE 00 2d 0f 00 NEW_ADDRS c2 33 c5 e4 Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 3a 30 00 00 00 20 11 d3 c8 90 55 62 22 90 55 14 1e ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 3a 31 00 00 00 20 11 ef 17 90 55 62 22 c2 33 c5 f0 ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 42 32 00 00 00 20 11 d1 c0 90 55 62 22 90 55 14 1e ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 42 33 00 00 00 20 11 ed 0f 90 55 62 22 c2 33 c5 f0 ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 3d 35 00 00 00 20 11 ce c5 90 55 62 22 90 55 14 1e ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 3d 36 00 00 00 20 11 ea 14 90 55 62 22 c2 33 c5 f0 ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 2c 37 00 40 00 20 06 b6 df 90 55 62 22 c2 06 b8 6e ... Incoming PPP Data on interface: slot:3/mod:1 IP_DATA 45 00 00 38 38 00 00 00 20 01 cb da 90 55 62 22 90 55 14 1e ... Incoming PPP Data on interface: slot:3/mod:1 UTCP_DATA 45 00 00 28 39 00 40 00 20 06 b4 e3 90 55 62 22 c2 06 b8 6e ... Incoming PPP Data on interface: slot:3/mod:1 6755 45 00 00 28 3a 00 40 00 20 06 b3 e3 90 55 62 22 c2 06 b8 6e ... Outgoing PPP Data on interface: slot:3/mod:1 CCP RESET_REQ 00 00 d5 58 06 a8 b2 1d 48 50 18 22 38 e1 c6 00 00 47 45 54 20 2f 69 6d 61 67 65 2e 6e 67 3b 73 70 61 63 65 64 65 73 63 3d 53 6d 69 74 68 42 61 72 6e 65 79 5f 49 6e 64 69 76 49 6e 76 65 73 74 6f 72 5f 34 36 38 78 36 30 5f 52 75 6e 4f 66 Incoming PPP Data on interface: slot:3/mod:1 CCP RESET_ACK 00 00 27 01 59 08 78 00 00 00 00 00 00 06 00 48 68 01 00 66 72 65 31 62 75 66 66 00 00 00 00 b3 b3 b3 b3 00 00 ff 03 c0 21 08 3f 00 1c 3c e1 d2 87 e3 19 4c 45 5b 0f df be ff 50 b7 82 e9 9e 95 a7 cd 47 27 a3 00 00 00 00 00 01 00 0a 00 00 Outgoing PPP Data on interface: slot:3/mod:1 CCP RESET_REQ 1d 58 00 c0 49 0d 8d dc 08 00 45 00 02 cc b2 03 40 00 1f 06 2c a1 90 55 62 0a c2 33 c5 f4 04 6d 1f 90 00 4b 71 40 7a 6d 75 e0 50 18 22 38 0a 89 00 00 47 45 54 20 68 74 74 70 3a 2f 2f 72 65 67 69 73 74 65 72 2e 6d 69 63 72 6f 73 6f 66 74 2e 63 6f 6d 2f 6c 69 62 72 61 72 79 2f 69 6d 61 67 65 73 2f 67 69 66 73 2f 74 6f 6f 6c 62 61 72 2f 6d 73 66 74 2e 67 69 66 20 48 54 54 50 2f 31 2e 30 0d 0a 52 65 66 65 72 65 72 3a 20 68 74 74 70 3a 2f 2f 72 65 67 69 73 74 65 72 2e 6d 69 63 72 6f 73 6f 66 74 2e 63 6f 6d 2f 52 45 47 57 49 5a 2f 66 6f 72 6d 73 Incoming PPP Data on interface: slot:3/mod:1 CCP RESET_ACK 1d 00 25 42 46 4c 52 00 00 00 00 00 00 00 80 00 01 a5 00 4d 45 4d 31 42 55 46 46 00 00 00 00 01 06 e1 4c 00 00 00 14 00 00 06 20 ff ff ff ff ff ff ff ff 08 00 00 00 00 00 06 3f 44 52 56 35 00 00 00 00 00 00 00 00 65 73 73 20 58 20 66 6f 72 20 68 65 78 20 74 72 61 63 69 6e 67 2e 0d 0a 00 41 41 41 41 41 41 41 41 41 41 41 41 41 41 41 41 41 41 00 00 21 00 01 00 a0 0f da 0b 32 ca 68 0d aa 20 0f 4d 50 54 53 00 00 00 00 00 00 00 40 00 01 a5 00 4d 45 4d 31 00 00 00 02 00 00 00 00 00 00 00 2a 00 62 bb 0c ce fb 06 9b 06 ae ff 03 00 2d d0 2b 50 10 22
Subject: Re: (usr-tc) PCSDL
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-11 10:53:31
Brian said once upon a time: > >Is it just me, or does PCSDL fail 75% of the time? I try it from NT >machines to reflash NMC code: Try it from DOS. The program is incredibly finicky. I tried using it on a Powerbook with SoftPC and it would crash the emulator. Whenever I use it from DOS on a real PC, it works fine.
Subject: Re: (usr-tc) Framed-Route, which one?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-11 10:56:08
Brian said once upon a time: >Which one of these would be correct? >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Netmask = 255.255.255.255, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP This one. >I am routing the /28 thru a single host of 208.214.45.1. I would think >the that one WITHOUT the Framed-Netmask, would tell the NAS > >208.214.44.1 255.255.255.255 >208.214.45.0 255.255.255.240 > >So do I need the Framed-Netmask, or not? Yes, but it only applies to the host route, not the subnet behind it.
Subject: Re: (usr-tc) Netserver route problem
From: Brian <signal@shreve.net>
Date: 1998-02-11 11:01:07
On Wed, 11 Feb 1998, Russ Panula wrote: > On Mon, 9 Feb 1998 15:19:27 -0600 (CST), Brian <signal@shreve.net> > wrote: > > >> >yet, whenever I start gated on a machine (running rip), I get this in the > >> >routing tables: > >> > > >> >208.214.45.0 usr2ts1.shreve. 255.255.255.0 UG 1500 0 eth0 > >> > >> Try adding > >> > >> Framed-Netmask = 255.255.255.255 > > > >Why would I delcare a Framed-Netmask on a Framed-Route account? > > Specifying a host netmask cleared up an erroneous route advertisement > I was seeing with hosts authenticated by similar RADIUS entries and > routing /28's. > > I'm guessing usr2ts1 is announcing a /24 route because the client in > question is being assigned inet 208.214.45.1 mask 255.255.255.0. > > What netmask is the client assigned? What routes are successfully > announced when the client connects? Framed-Route = "208.214.45.0/28 208.214.45.1 1" above declares a netmask of 255.255.255.240, does it not? AFAIK, you should not *need* Framed-Netmasks, when using Framed-Routes, but I would really like someone to clear this up. I was under the impression that you didn't, and if you do, it could be a bug. Other networks route fine with the above statment, its only the .0 networks where I see this anomaly. I will try the framed netmask of 255.255.255.255, and see how that goes. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Hiper ISDN problems
From: Brian <signal@shreve.net>
Date: 1998-02-11 11:02:05
On Wed, 11 Feb 1998, Robert von Bismarck wrote: > We have several customers with this problem, they connect just fine with > the Total Control and the HiPerDSP cards, but are unable to surf, I > traced several with the monitoring commands of the HiPerARC, and found > some weird things. The interface asks for quite a lot of RESET's. Here's > an example, if anyone has seen something like it and got a solution, > please let me know. > The customers are using Win95 without DUN 1.2, with a config that worked > often. > > Thanks for any info, > > Robert von Bismarck > Petrel Communication S.A. I just fixed a bunch of these by having them download the latest DUN and TCP/IP upgrades at microsoft.com, there are several even for 95b Brian > > > Outgoing PPP Data on interface: slot:3/mod:1 > CHAP SUCCESS 00 > > Outgoing PPP Data on interface: slot:3/mod:1 > IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS c2 33 c5 e4 > > Incoming PPP Data on interface: slot:3/mod:1 > CCP CFG_REQ STAC_COMP 00 01 03 > > Outgoing PPP Data on interface: slot:3/mod:1 > CCP CFG_ACK STAC_COMP 00 01 03 > > Outgoing PPP Data on interface: slot:3/mod:1 > CCP CFG_REQ STAC_COMP 00 01 03 > > Incoming PPP Data on interface: slot:3/mod:1 > CCP CFG_ACK STAC_COMP 00 01 03 > > Incoming PPP Data on interface: slot:3/mod:1 > IPCP CFG_REQ COMPR_TYPE 00 2d 0f 01 > NEW_ADDRS 00 00 00 00 > 81 00 00 00 00 > 82 00 00 00 00 > 83 00 00 00 00 > 84 00 00 00 00 > > Outgoing PPP Data on interface: slot:3/mod:1 > IPCP CFG_REJ 82 00 00 00 00 > 83 00 00 00 00 > 84 00 00 00 00 > > Incoming PPP Data on interface: slot:3/mod:1 > IPCP CFG_REQ COMPR_TYPE 00 2d 0f 01 > NEW_ADDRS 00 00 00 00 > 81 00 00 00 00 > > Outgoing PPP Data on interface: slot:3/mod:1 > IPCP CFG_NAK COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS 90 55 62 22 > 81 c2 33 c5 f0 > > Incoming PPP Data on interface: slot:3/mod:1 > IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS 90 55 62 22 > 81 c2 33 c5 f0 > > Outgoing PPP Data on interface: slot:3/mod:1 > IPCP CFG_ACK COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS 90 55 62 22 > 81 c2 33 c5 f0 > > Outgoing PPP Data on interface: slot:3/mod:1 > IPCP CFG_REQ COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS c2 33 c5 e4 > > Incoming PPP Data on interface: slot:3/mod:1 > IPCP CFG_ACK COMPR_TYPE 00 2d 0f 00 > NEW_ADDRS c2 33 c5 e4 > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 3a 30 00 00 00 20 11 d3 c8 90 55 62 22 90 55 14 > 1e ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 3a 31 00 00 00 20 11 ef 17 90 55 62 22 c2 33 c5 > f0 ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 42 32 00 00 00 20 11 d1 c0 90 55 62 22 90 55 14 > 1e ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 42 33 00 00 00 20 11 ed 0f 90 55 62 22 c2 33 c5 > f0 ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 3d 35 00 00 00 20 11 ce c5 90 55 62 22 90 55 14 > 1e ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 3d 36 00 00 00 20 11 ea 14 90 55 62 22 c2 33 c5 > f0 ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 2c 37 00 40 00 20 06 b6 df 90 55 62 22 c2 06 b8 > 6e ... > > Incoming PPP Data on interface: slot:3/mod:1 > IP_DATA 45 00 00 38 38 00 00 00 20 01 cb da 90 55 62 22 90 55 14 > 1e ... > > Incoming PPP Data on interface: slot:3/mod:1 > UTCP_DATA 45 00 00 28 39 00 40 00 20 06 b4 e3 90 55 62 22 c2 06 b8 > 6e ... > > Incoming PPP Data on interface: slot:3/mod:1 > 6755 45 00 00 28 3a 00 40 00 20 06 b3 e3 90 55 62 22 c2 06 b8 > 6e ... > > Outgoing PPP Data on interface: slot:3/mod:1 > CCP RESET_REQ 00 00 d5 58 06 a8 b2 1d 48 > 50 18 22 38 e1 c6 00 00 > 47 45 54 20 2f 69 6d 61 > 67 65 2e 6e 67 3b 73 70 > 61 63 65 64 65 73 63 3d > 53 6d 69 74 68 42 61 72 > 6e 65 79 5f 49 6e 64 69 > 76 49 6e 76 65 73 74 6f > 72 5f 34 36 38 78 36 30 > 5f 52 75 6e 4f 66 > > Incoming PPP Data on interface: slot:3/mod:1 > CCP RESET_ACK 00 00 27 01 59 08 78 00 00 > 00 00 00 00 06 00 48 68 > 01 00 66 72 65 31 62 75 > 66 66 00 00 00 00 b3 b3 > b3 b3 00 00 ff 03 c0 21 > 08 3f 00 1c 3c e1 d2 87 > e3 19 4c 45 5b 0f df be > ff 50 b7 82 e9 9e 95 a7 > cd 47 27 a3 00 00 00 00 > 00 01 00 0a 00 00 > > Outgoing PPP Data on interface: slot:3/mod:1 > CCP RESET_REQ 1d 58 00 c0 49 0d 8d dc 08 > 00 45 00 02 cc b2 03 40 > 00 1f 06 2c a1 90 55 62 > 0a c2 33 c5 f4 04 6d 1f > 90 00 4b 71 40 7a 6d 75 > e0 50 18 22 38 0a 89 00 > 00 47 45 54 20 68 74 74 > 70 3a 2f 2f 72 65 67 69 > 73 74 65 72 2e 6d 69 63 > 72 6f 73 6f 66 74 2e 63 > 6f 6d 2f 6c 69 62 72 61 > 72 79 2f 69 6d 61 67 65 > 73 2f 67 69 66 73 2f 74 > 6f 6f 6c 62 61 72 2f 6d > 73 66 74 2e 67 69 66 20 > 48 54 54 50 2f 31 2e 30 > 0d 0a 52 65 66 65 72 65 > 72 3a 20 68 74 74 70 3a > 2f 2f 72 65 67 69 73 74 > 65 72 2e 6d 69 63 72 6f > 73 6f 66 74 2e 63 6f 6d > 2f 52 45 47 57 49 5a 2f > 66 6f 72 6d 73 > > Incoming PPP Data on interface: slot:3/mod:1 > CCP RESET_ACK 1d 00 25 42 46 4c 52 00 00 > 00 00 00 00 00 80 00 01 > a5 00 4d 45 4d 31 42 55 > 46 46 00 00 00 00 01 06 > e1 4c 00 00 00 14 00 00 > 06 20 ff ff ff ff ff ff > ff ff 08 00 00 00 00 00 > 06 3f 44 52 56 35 00 00 > 00 00 00 00 00 00 65 73 > 73 20 58 20 66 6f 72 20 > 68 65 78 20 74 72 61 63 > 69 6e 67 2e 0d 0a 00 41 > 41 41 41 41 41 41 41 41 > 41 41 41 41 41 41 41 41 > 41 00 00 21 00 01 00 a0 > 0f da 0b 32 ca 68 0d aa > 20 0f 4d 50 54 53 00 00 > 00 00 00 00 00 40 00 01 > a5 00 4d 45 4d 31 00 00 > 00 02 00 00 00 00 00 00 > 00 2a 00 62 bb 0c ce fb > 06 9b 06 ae ff 03 00 2d > d0 2b 50 10 22 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Framed-Route, which one?
From: Brian <signal@shreve.net>
Date: 1998-02-11 11:08:53
Which one of these would be correct? signal Authentication-Type = Unix-PW Service-Type = Framed, Framed-Protocol = PPP, Framed-IP-Address = 208.214.45.1, Framed-Route = "208.214.45.0/28 208.214.45.1 1", Framed-IP-Netmask = 255.255.255.240, Framed-MTU = 1500, Framed-Compression = Van-Jacobson-TCP-IP or signal Authentication-Type = Unix-PW Service-Type = Framed, Framed-Protocol = PPP, Framed-IP-Address = 208.214.45.1, Framed-Netmask = 255.255.255.255, Framed-Route = "208.214.45.0/28 208.214.45.1 1", Framed-IP-Netmask = 255.255.255.240, Framed-MTU = 1500, Framed-Compression = Van-Jacobson-TCP-IP I am routing the /28 thru a single host of 208.214.45.1. I would think the that one WITHOUT the Framed-Netmask, would tell the NAS 208.214.44.1 255.255.255.255 208.214.45.0 255.255.255.240 So do I need the Framed-Netmask, or not? /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NMC card code
From: jason_kelton@3com.com
Date: 1998-02-11 11:11:32
Yeah... I think its looking for the NMC 16MEG Flash version of code. 5.2.2 is the current version you should be looking for. Regards, Jason Kelton Carrier Systems Engineer 3Com ANZA. jeffm@iglou.com on 11/02/98 08:45:31 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) I have an NMC card in a fairly new chassis that, when pulled up in TCM, is asking for hm*.nac code to upgrade it. I can't find this code anywhere (TCM seems to be label it as NMC code (8 MB) or something like that. Any idea what this thing is asking for? -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456 - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: (usr-tc) PCSDL
From: Brian <signal@shreve.net>
Date: 1998-02-11 11:29:52
Is it just me, or does PCSDL fail 75% of the time? I try it from NT machines to reflash NMC code: NMC card, dip switches 1 & 2 set to on (57600) NT: com2 set for 57600, hardware flow control, fifo enabled I usually have to try the pcsdl a few times before it works. I erase the flash of the nmc before I start the pcsdl. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-11 11:50:16
On Wed, 11 Feb 1998, Greg Coffey wrote: > According to Eric Benhamou, the head honcho at 3Com, "The V.pcm standard is > so close to x2 that 3Com expects to ship V.pcm-compliant products within > days after the standard is voted on. "We'll provide a free [downloadable] > software upagrade to our x2 customers who want to run V.pcm. As it stands > today, the content of standard has been resolved, and the confididence of > having a formal V.pcm modem standard is extremely high." Doesn't the ITU meet and vote in September, which means V.PCM won't be shipped until then. The web site at http://www.3com.com/56k says that upgrades to the standard will be available for download on February 16th.
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-11 12:11:50
At 10:48 AM 2/10/98 -0600, Tatai SV Krishnan wrote: >Check it out on www.3com.com > >krish I did.. But it did not say anything about Total Controls, I think it only refered to modems, so I wanted to make sure ;) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) NMC card code
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-11 12:16:48
We have 16MB of RAM for our NMC cards, but we only have 2MB flash. Does the new code actually require more flash memory to work? The SDL image is only 1.5 MB. Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- Cc: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> On Tue, 10 Feb 1998, Curt Shambeau wrote: > > I have an NMC card in a fairly new chassis that, when pulled up in TCM, > > is asking for hm*.nac code to upgrade it. I can't find this code > > anywhere (TCM seems to be label it as NMC code (8 MB) or something like > > that. Any idea what this thing is asking for? > > Any NMC code that starts with HM is for the Hiper Chassis. It means the > NMC has already been upgraded with RAM and a larger flash ROM. >Note that 3COM has been shipping the NMC cards with 16MB ram and 8MB (Or >is it 4MB?) flash for some time now. > >I was suprised to see that my 2059 chassises purchased in Sept and Dec >already have the NMC upgrades. > >Brian
Subject: Re: (usr-tc) NMC card code
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-11 12:22:24
> We have 16MB of RAM for our NMC cards, but we only have 2MB flash. Does the > new code actually require more flash memory to work? The SDL image is only > 1.5 MB. My accident, we once flashed the new NMC code for a 16MB box into a 4MB box. It sorta worked, so maybe it was just a memory thing instead of a flash thing. Can't hurt to try - worst case you can use PCSDL to flash the right code in. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Framed-Route, which one?
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-11 12:28:01
In a case like this we usually just set: Framed-Address = 208.214.45.1 Framed-Netmask = 255.255.255.240 We don't usually use the Framed-Route, Framed-IP-Address, or the Framed-IP-Netmask attributes. When it brings up the interface with the framed address and netmask, it automatically creates a route as 208.214.45.0/28 208.214.45.1. It's much easier to deal with this way. Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- > > >Which one of these would be correct? > >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > >or > >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Netmask = 255.255.255.255, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > > >I am routing the /28 thru a single host of 208.214.45.1. I would think >the that one WITHOUT the Framed-Netmask, would tell the NAS > >208.214.44.1 255.255.255.255 >208.214.45.0 255.255.255.240 > >So do I need the Framed-Netmask, or not? > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Network Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | >\-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NMC card code
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-11 12:44:37
On Wed, 11 Feb 1998, Curt Shambeau wrote: > > We have 16MB of RAM for our NMC cards, but we only have 2MB flash. > Does the > new code actually require more flash memory to work? The SDL > image is only > 1.5 MB. > My accident, we once flashed the new NMC code for a 16MB box into a 4MB > box. It sorta worked, so maybe it was just a memory thing instead of a > flash thing. Can't hurt to try - worst case you can use PCSDL to flash > the right code in. Are there any significant differences between the 16MB & 4MB versions of the NMC software? Is it worth the upgrade?
Subject: (usr-tc) How do you get PRI ISDN calls to terminate in the modem, rather than , the netserver?
From: Jaye Mathisen <mrcpu@cdsnet.net>
Date: 1998-02-11 12:47:43
-
Subject: Re: (usr-tc) Framed-Route, which one?
From: Brian <signal@shreve.net>
Date: 1998-02-11 12:52:14
On Wed, 11 Feb 1998, Peter D. Mayer wrote: > In a case like this we usually just set: > > Framed-Address = 208.214.45.1 > Framed-Netmask = 255.255.255.240 > > We don't usually use the Framed-Route, Framed-IP-Address, or the > Framed-IP-Netmask attributes. When it brings up the interface with the > framed address and netmask, it automatically creates a route as > 208.214.45.0/28 208.214.45.1. It's much easier to deal with this way. Framed-Route = "208.214.45.0/28 208.214.45.1 1", seems easier to me. Thats why I am hoping it works, and Framed-Route I think is the way its suppose to be done. Brian > > Peter D. Mayer > NetWalk Tech Support > dmayer@netwalk.com > > -----Original Message----- > From: Brian <signal@shreve.net> > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Date: Wednesday, February 11, 1998 12:11 PM > Subject: (usr-tc) Framed-Route, which one? > > > > > > > >Which one of these would be correct? > > > >signal Authentication-Type = Unix-PW > > Service-Type = Framed, > > Framed-Protocol = PPP, > > Framed-IP-Address = 208.214.45.1, > > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > > Framed-IP-Netmask = 255.255.255.240, > > Framed-MTU = 1500, > > Framed-Compression = Van-Jacobson-TCP-IP > > > >or > > > >signal Authentication-Type = Unix-PW > > Service-Type = Framed, > > Framed-Protocol = PPP, > > Framed-IP-Address = 208.214.45.1, > > Framed-Netmask = 255.255.255.255, > > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > > Framed-IP-Netmask = 255.255.255.240, > > Framed-MTU = 1500, > > Framed-Compression = Van-Jacobson-TCP-IP > > > > > >I am routing the /28 thru a single host of 208.214.45.1. I would think > >the that one WITHOUT the Framed-Netmask, would tell the NAS > > > >208.214.44.1 255.255.255.255 > >208.214.45.0 255.255.255.240 > > > >So do I need the Framed-Netmask, or not? > > > >/-------------------------- > signal@shreve.net -----------------------------\ > >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 > | > >| Network Administrator | Perl, Linux | Web hosting, online stores, > | > >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs > | > >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ > | > >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 > | > >\-------------------------- 318-222-2638 > x109 -----------------------------/ > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) PCSDL
From: Brian <signal@shreve.net>
Date: 1998-02-11 12:53:48
On Wed, 11 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > > >Is it just me, or does PCSDL fail 75% of the time? I try it from NT > >machines to reflash NMC code: > > Try it from DOS. The program is incredibly finicky. I tried using it on a > Powerbook with SoftPC and it would crash the emulator. Whenever I use it > from DOS on a real PC, it works fine. Everything here is Win95 or NT, that would run pcsdl. I would LOVE a pcsdl utility for Linux, and I bet I could write one if I had source code to windows pcsdl, which is never going to happen. I will try a laptop booted in "dos mode" Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Brian <signal@shreve.net>
Date: 1998-02-11 12:54:33
On Wed, 11 Feb 1998, Lee Kuo wrote: > On Wed, 11 Feb 1998, Greg Coffey wrote: > > According to Eric Benhamou, the head honcho at 3Com, "The V.pcm standard is > > so close to x2 that 3Com expects to ship V.pcm-compliant products within > > days after the standard is voted on. "We'll provide a free [downloadable] > > software upagrade to our x2 customers who want to run V.pcm. As it stands > > today, the content of standard has been resolved, and the confididence of > > having a formal V.pcm modem standard is extremely high." > > Doesn't the ITU meet and vote in September, which means V.PCM won't be > shipped until then. The web site at http://www.3com.com/56k says that > upgrades to the standard will be available for download on February 16th. Yes, the ITU meets in September, so it won't be adopted as an "official" ITU standard until they meet and vote, but vendors can still ship it. Brian > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Framed-Route, which one?
From: Brian <signal@shreve.net>
Date: 1998-02-11 12:56:08
On Wed, 11 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > >Which one of these would be correct? > > >signal Authentication-Type = Unix-PW > > Service-Type = Framed, > > Framed-Protocol = PPP, > > Framed-IP-Address = 208.214.45.1, > > Framed-Netmask = 255.255.255.255, > > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > > Framed-IP-Netmask = 255.255.255.240, > > Framed-MTU = 1500, > > Framed-Compression = Van-Jacobson-TCP-IP > > This one. > > >I am routing the /28 thru a single host of 208.214.45.1. I would think > >the that one WITHOUT the Framed-Netmask, would tell the NAS > > > >208.214.44.1 255.255.255.255 > >208.214.45.0 255.255.255.240 > > > >So do I need the Framed-Netmask, or not? > > Yes, but it only applies to the host route, not the subnet behind it. ok so use Framed-Netmask always then, gotcha. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-11 12:58:08
Thus spake Lee Kuo >On Wed, 11 Feb 1998, Greg Coffey wrote: >> According to Eric Benhamou, the head honcho at 3Com, "The V.pcm standard is >> so close to x2 that 3Com expects to ship V.pcm-compliant products within >> days after the standard is voted on. "We'll provide a free [downloadable] >> software upagrade to our x2 customers who want to run V.pcm. As it stands >> today, the content of standard has been resolved, and the confididence of >> having a formal V.pcm modem standard is extremely high." >Doesn't the ITU meet and vote in September, which means V.PCM won't be >shipped until then. The web site at http://www.3com.com/56k says that >upgrades to the standard will be available for download on February 16th. That's for final ratifications...draft standard will be (or has been) voted on in Feb. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re[2]: (usr-tc) ITU Standard
From: phenkle@usr.com
Date: 1998-02-11 12:59:28
--IMA.Boundary.127322788 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part All x2 customers (Sportsters, Netservers, Total Control, etc.) are entitled to a free v.90 upgrade. (after all it is just software) However, Total Control will auto-negotiate x2 or v.90 calls, so the client modems do not have to change :) I hope this helps Pat ______________________________ Reply Separator _________________________________ Author: Ken Ogura <ogura@telcel.net.ve> at Internet Or to a new standard, V.90 Terry Kennedy wrote: > So when USR (3com) puts out the upgrade to the > new standard (v.pcm?) will all the customers we have > that have bought sporters or other X2 modems need to upgrade > their modems? This could be real nightmare... > > Or will be so cool as to still work with X2 products? > After reading the white papers at 3com.com it was still unclear > to me. Any clarifications would help. Thanks... > > Terry Kennedy > System Admin OlyPen, Inc. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. --IMA.Boundary.127322788 Content-Type: text/plain; charset=US-ASCII; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Ken Ogura n: Ogura;Ken org: Telcel adr: Av. Francisco de Miranda, Parque Canaima;;Piso 16, Los Palos Grandes;Caracas;Miranda;;Venezuela email;internet: ogura@telcel.net.ve title: Gerente Operaciones T-Net tel;work: 58-2-2018170 tel;fax: 58-2-2018140 tel;home: 58-14-149922 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --IMA.Boundary.127322788 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4E1E7FA0; Wed, 11 Feb 98 12:03:38 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id LAA12988; Wed, 11 Feb 1998 11:44:32 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0y2gQy-0007RX-00; Wed, 11 Feb 1998 10:58:56 -0700 Received: from smarty.telcel.net.ve [206.48.41.100] by lists.xmission.com with esmtp (Exim 1.73 #4) id 0y2gQu-0007Ib-00; Wed, 11 Feb 1998 10:58:52 -0700 Received: from telcel.net.ve ([172.16.3.133]) by smarty.telcel.net.ve (Post.Office MTA v3.1 release 0154 ID# 557-42303U60000L60000S0) with ESMTP id AAA17555 for <usr-tc@lists.xmission.com>; Wed, 11 Feb 1998 13:53:07 -0400 Message-ID: <34E1E43A.47758AE9@telcel.net.ve> Organization: Telcel X-Mailer: Mozilla 4.04 [en] (Win95; I) MIME-Version: 1.0 References: <m0y2gLq-0005GcC@mail.olypen.com> Content-Type: multipart/mixed; boundary="------------C650A42C0DFB770BF188028C" Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.127322788--
Subject: Re: (usr-tc) NMC card code
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-11 13:10:51
> Are there any significant differences between the 16MB & 4MB versions of > the NMC software? Is it worth the upgrade? One can manage HiPer equipment (16mb version), and the other can not. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) ITU Standard
From: Ken Ogura <ogura@telcel.net.ve>
Date: 1998-02-11 13:47:39
This is a multi-part message in MIME format. --------------C650A42C0DFB770BF188028C Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Or to a new standard, V.90 Terry Kennedy wrote: > So when USR (3com) puts out the upgrade to the > new standard (v.pcm?) will all the customers we have > that have bought sporters or other X2 modems need to upgrade > their modems? This could be real nightmare... > > Or will be so cool as to still work with X2 products? > After reading the white papers at 3com.com it was still unclear > to me. Any clarifications would help. Thanks... > > Terry Kennedy > System Admin OlyPen, Inc. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. --------------C650A42C0DFB770BF188028C Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Ogura, Ken Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Ken Ogura n: Ogura;Ken org: Telcel adr: Av. Francisco de Miranda, Parque Canaima;;Piso 16, Los Palos Grandes;Caracas;Miranda;;Venezuela email;internet: ogura@telcel.net.ve title: Gerente Operaciones T-Net tel;work: 58-2-2018170 tel;fax: 58-2-2018140 tel;home: 58-14-149922 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------C650A42C0DFB770BF188028C--
Subject: Re: (usr-tc) RADIUS, PPP, disconnect
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-11 14:05:59
yong_chen@nyed.uscourts.gov said once upon a time: >The problem that is only one user(doesn't matter which one) in Merit RADIUS >server user table can login to the system and stay in. None of the other users >in that user table can login and stay in(they got kick out right after the Merit >RADIUS server verify their password). >We already had try three different version of Merit RADIUS servers in Solaris. >Why only one user can login? Is this is a known problem? Please Help. It sounds like you have some problems with your "users" file. Show me what the defaults and the one good user look like. Then I'd like to see an example of a bad user entry.
Subject: Re: Re[2]: (usr-tc) ITU Standard
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-11 14:16:02
On Wed, 11 Feb 1998 phenkle@usr.com wrote: > All x2 customers (Sportsters, Netservers, Total Control, etc.) are > entitled to a free v.90 upgrade. (after all it is just software) > > However, Total Control will auto-negotiate x2 or v.90 calls, so the > client modems do not have to change :) Is the Total Control modem code going to be out before the Sportster/Courier modem code this time? When x2 was released, the client code was available a week or two in advance, and the ISPs like us, who signed up in advance on the x2 web site, were pummeled with users asking when we were going to have x2 available since their modems supported it. Will client modems with V.90 be able to drop back to x2 if the server modem supports only x2 and not V.90? Brian
Subject: Re: (usr-tc) How do you get PRI ISDN calls to terminate in the modem, rather than , the netserver?
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-11 14:54:28
Jaye Mathisen said once upon a time: > >How do you get PRI ISDN calls to terminate in the modem, rather than , >the netserver? Set ISDN GW Slot to 0 in the PRI card configuration.
Subject: (usr-tc) RADIUS, PPP, disconnect
From: yong_chen@nyed.uscourts.gov
Date: 1998-02-11 15:30:53
We have a Total Control Hub with dual Pri card, NetServer card and quad modems. We also install Merit RADIUS server to prevent unauthurize access. Users who use PPP login to our system WITHOUT checked by the Merit RADIUS server are NO problem. The problem that is only one user(doesn't matter which one) in Merit RADIUS server user table can login to the system and stay in. None of the other users in that user table can login and stay in(they got kick out right after the Merit RADIUS server verify their password). We already had try three different version of Merit RADIUS servers in Solaris. Why only one user can login? Is this is a known problem? Please Help. Yong U.S District Courts, Eastern district of NY 718-260-2297 fax 718-260-2338
Subject: Re: (usr-tc) ITU Standard
From: Greg Coffey <greg@coffey.com>
Date: 1998-02-11 16:08:42
Yep, I thought that little sleight of hand was a poor way of doing business. I don't know how many angry customers I had to two-step that issue with. Maybe 3Com does not define FREE as a $60 charge. At 06:29 PM 2/11/98 -0400, you wrote: >It=B4s really free or USR will charge 60$ for upgrade. Remenber V.34 to X2 >"Upgradeble"!!! > >Brian Elfert wrote: > >> On Wed, 11 Feb 1998 phenkle@usr.com wrote: >> >> > All x2 customers (Sportsters, Netservers, Total Control, etc.) are >> > entitled to a free v.90 upgrade. (after all it is just software) >> > >> > However, Total Control will auto-negotiate x2 or v.90 calls, so= the >> > client modems do not have to change :) Thanks, Greg Coffey, CoffeyNet ** $20 local Casper USR x2 56k access ** =20 142 S. Center St. Wheatland, Pinedale, Lander, Lusk Casper, WY 82601 Douglas & Rawlins (307) 234-5443 http://www.coffey.com Open 8-6 M-F / 10-2 Saturday
Subject: Re: (usr-tc) New Netserver code out yesterday
From: jcondon <jcondon@computer.net>
Date: 1998-02-11 16:23:25
> >Doesn't the ITU meet and vote in September, which means V.PCM won't be > >shipped until then. The web site at http://www.3com.com/56k says that > >upgrades to the standard will be available for download on February 16th. > > That's for final ratifications...draft standard will be (or has been) > voted on in Feb. Also the URL above says NOTHING about when the code will be out. It only says check here on or after the 16th for when it will be available. So it may or may not be available on the 16th the website didn't give a solid date.
Subject: Re: (usr-tc) Framed-Route, which one?
From: Russell Heilling <russellh@netdirect.net.uk>
Date: 1998-02-11 17:58:17
Framed-Netmask is actually the same attribute as Framed-IP-Netmask, it is just there for backwards compatibility with Radius 1 user entries. To do what you are trying here I would use: signal Auth-Type=System Service-Type=Framed-User, Framed-Protocol=PPP, Framed-IP-Address=208.214.45.1, Framed-IP-Netmask=255.255.255.240, Framed-MTU=1500, Framed-Compression=Van-Jacobson-TCP-IP, The Framed-Route would be needed if you wanted to give the user an address outside of his subnet for the ISDN side of the router and a subnet for the ethernet side, and you could route the subnet through his dial-up IP address. This is not needed in most routers, as you can run the ISDN as an unnumbered PPP link. --- Russell Heilling -- NetDirect Internet Ltd Tel: +44 (0) 181 293 7000 Fax: +44 (0) 181 293 6000 email: russellh@netdirect.net.uk -----Original Message----- > > >Which one of these would be correct? > >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > >or > >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Netmask = 255.255.255.255, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP > > >I am routing the /28 thru a single host of 208.214.45.1. I would think >the that one WITHOUT the Framed-Netmask, would tell the NAS > >208.214.44.1 255.255.255.255 >208.214.45.0 255.255.255.240 > >So do I need the Framed-Netmask, or not? > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Network Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) ITU Standard
From: Ken Ogura <ogura@telcel.net.ve>
Date: 1998-02-11 18:29:36
This is a multi-part message in MIME format. --------------7012D682B96D34228FD6E16A Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit It�s really free or USR will charge 60$ for upgrade. Remenber V.34 to X2 "Upgradeble"!!! Brian Elfert wrote: > On Wed, 11 Feb 1998 phenkle@usr.com wrote: > > > All x2 customers (Sportsters, Netservers, Total Control, etc.) are > > entitled to a free v.90 upgrade. (after all it is just software) > > > > However, Total Control will auto-negotiate x2 or v.90 calls, so the > > client modems do not have to change :) > > Is the Total Control modem code going to be out before the > Sportster/Courier modem code this time? > > When x2 was released, the client code was available a week or two in > advance, and the ISPs like us, who signed up in advance on the x2 web > site, were pummeled with users asking when we were going to have x2 > available since their modems supported it. > > Will client modems with V.90 be able to drop back to x2 if the server > modem supports only x2 and not V.90? > > Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. --------------7012D682B96D34228FD6E16A Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Ogura, Ken Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Ken Ogura n: Ogura;Ken org: Telcel adr: Av. Francisco de Miranda, Parque Canaima;;Piso 16, Los Palos Grandes;Caracas;Miranda;;Venezuela email;internet: ogura@telcel.net.ve title: Gerente Operaciones T-Net tel;work: 58-2-2018170 tel;fax: 58-2-2018140 tel;home: 58-14-149922 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------7012D682B96D34228FD6E16A--
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-11 21:11:35
> > Are you putting the quad code on chassis that do not have the new > > netserver code? Is it safe to put the new quad code on 3.6.x chassis? > > No, I'm pretty sure they do no recommend running the new Quad code on the > older 3.6.28 netserver code. OK, that brings up the opposite question - is it safe to run 3.7.x with the OLD (5.5.7/5.6.7) Quad firmware? We prefer to change one thing at a time here, so if something breaks we know where to point the finger. We typically do the NETServer first, followed by the Quads a week or two later. We only do the PRI cards if it's required since if the upgrade fails I can't get at the console remotely (yet) :-( Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: Re: (usr-tc) New Netserver code out yesterday
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-11 21:13:46
> I'm wondering why exactly this new modem code requires new Netserver code? > > The old code revs have never need new Netserver code. Maybe a change in the way PPP-in-modem works? Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: Re: (usr-tc) DNIS call map to service
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-11 22:34:44
On Thu, 12 Feb 1998, Douglas Palmer wrote: > This is probably fairly elementary, but I can't locate the answer in the docs. > > I want a certain called number to automatically telnet to a certain address. > > For example, if a user dials into 718-555-4429, I want that user to > automatically telnet to 156.121.32.227. No authentication, no other > services -- just telnet to that one address. This should be possible with > DNIS call mapping, right? Any help would be appreciated. > > Thanks for your help. > DCP You can do this with Radius. You however have to modify the radius to do this. If you are using USR radius then the modification is just in the radser.scp script file. If you are using livingston Radius then you can modify the radiusd.h radiusd.c and users file. I have this done in 1.16 livingston release. I can tell you what to do if you need this info. krish > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: (usr-tc) HiPer DSP cards
From: Matthew Opoka <phantom@magnolia.net>
Date: 1998-02-11 23:23:43
I just got my HiPer DSP cards in with the memory upgrade. Do I need the memory upgrade? Which card is the memory & flash suppose to go in? My NMC already has 16MB RAM and 8MB flash. My Netserver has 16MB RAM and 4MB flash.
Subject: Re: (usr-tc) Framed-Route, which one?
From: MegaZone <megazone@megazone.org>
Date: 1998-02-12 00:00:21
Once upon a time Brian shaped the electrons to say... >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP This looks ok. >signal Authentication-Type = Unix-PW > Service-Type = Framed, > Framed-Protocol = PPP, > Framed-IP-Address = 208.214.45.1, > Framed-Netmask = 255.255.255.255, > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > Framed-IP-Netmask = 255.255.255.240, > Framed-MTU = 1500, > Framed-Compression = Van-Jacobson-TCP-IP This is bogus. Framed-Netmask is the old name for Framed-IP-Netmask, so you are seting the same thing twice to different values. -MZ -- <URL:mailto:megazone@livingston.com> Gweep, author, webmaster, human being, me <URL:mailto:megazone@megazone.org> H:510-527-0944 W:800-458-9966 510-426-0770 <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia!
Subject: Re: (usr-tc) New Netserver code out yesterday
From: MegaZone <megazone@megazone.org>
Date: 1998-02-12 00:02:04
Once upon a time Lee Kuo shaped the electrons to say... >Doesn't the ITU meet and vote in September, which means V.PCM won't be >shipped until then. The web site at http://www.3com.com/56k says that >upgrades to the standard will be available for download on February 16th. 1. It is V.90, not V.PCM - V.PCM was the working name. 2. The WG agreed on the preliminary V.90, the hope and plan is that it will just be rubber stamped as official in September. But it is officially an ITU-T proposed standard at this point after the February meeting. -MZ -- <URL:mailto:megazone@livingston.com> Gweep, author, webmaster, human being, me <URL:mailto:megazone@megazone.org> H:510-527-0944 W:800-458-9966 510-426-0770 <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia!
Subject: Re: (usr-tc) PCSDL
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-12 06:06:04
You should be able to run it with dosemu under Linux, since all pcsdl.exe does is access the serial port.. - lv On Wed, 11 Feb 1998, Brian wrote: > On Wed, 11 Feb 1998, Pete Ashdown wrote: > > > Brian said once upon a time: > > > > > >Is it just me, or does PCSDL fail 75% of the time? I try it from NT > > >machines to reflash NMC code: > > > > Try it from DOS. The program is incredibly finicky. I tried using it on a > > Powerbook with SoftPC and it would crash the emulator. Whenever I use it > > from DOS on a real PC, it works fine. > > Everything here is Win95 or NT, that would run pcsdl. I would LOVE a > pcsdl utility for Linux, and I bet I could write one if I had source code > to windows pcsdl, which is never going to happen. > > I will try a laptop booted in "dos mode" > > Brian > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) HiPer DSP cards
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-12 08:46:19
> I just got my HiPer DSP cards in with the memory upgrade. > Do I need the memory upgrade? > Which card is the memory & flash suppose to go in? > My NMC already has 16MB RAM and 8MB flash. > My Netserver has 16MB RAM and 4MB flash. The memory upgrade is meant for the NMC card. However, since you already have an upgraded card, you can just hold onto it as spare equipment. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: RE: (usr-tc) How do you get PRI ISDN calls to terminate in the modem, rather than , the netserver?
From: Sergio Manzi <smz@gpnet.it>
Date: 1998-02-12 08:48:42
On Wednesday, February 11, 1998 22:54, Pete Ashdown [SMTP:pashdown@xmission.com] wrote: > Jaye Mathisen said once upon a time: > > > >How do you get PRI ISDN calls to terminate in the modem, rather than , > >the netserver? > > Set ISDN GW Slot to 0 in the PRI card configuration. Do you mean that it is feasible to to route ISDN (I mean true ISDN, digital, PPP encapsulated calls) to the quad modem and from that to the async port on the NIC? We've been asking for that for months and our USR sales rep always told us that digital calls MUST terminate to the netserver/HyperArc.... We have 4 racks which are now populated with quads (analog/digital) and Network Management Card. Currently they are connected to the POTS network and to cisco 2511 routers via NIC. We would like to continue to use the 2511s but would like to get rid of the POTS, connect to PRI and be able to handle both analog and digital calls. Is that a dream? Asking for more: would it be feasible to have some quads receiving calls from the POTS (NIC) and others from the PRI card con the same chassis? This would be an even nicer dream for the transition. Cheers Sergio Manzi GP Net S.r.l.
Subject: Re: (usr-tc) HiPer DSP cards
From: Brian <signal@shreve.net>
Date: 1998-02-12 08:59:01
On Wed, 11 Feb 1998, Matthew Opoka wrote: > I just got my HiPer DSP cards in with the memory upgrade. > Do I need the memory upgrade? Yes thats why they sent it. Unless your chassis is a newer 2509 chassis that already has the memory upgrade. Check the DRAM size of your NMC via TCM. > Which card is the memory & flash suppose to go in? The NMC just like the instructions said. > My NMC already has 16MB RAM and 8MB flash. Then you can save the upgrade for another chassis > My Netserver has 16MB RAM and 4MB flash. The upgrade is for the NMC. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: RE: (usr-tc) How do you get PRI ISDN calls to terminate in the modem, rather than , the netserver?
From: Brian <signal@shreve.net>
Date: 1998-02-12 09:01:27
On Thu, 12 Feb 1998, Sergio Manzi wrote: > On Wednesday, February 11, 1998 22:54, Pete Ashdown > [SMTP:pashdown@xmission.com] wrote: > > Jaye Mathisen said once upon a time: > > > > > >How do you get PRI ISDN calls to terminate in the modem, rather than , > > >the netserver? > > > > Set ISDN GW Slot to 0 in the PRI card configuration. > > Do you mean that it is feasible to to route ISDN (I mean true ISDN, > digital, PPP encapsulated calls) to the quad modem and from that to the > async port on the NIC? We've been asking for that for months and our USR the nics not involved. You go from the PRI card, to the Modem NAC, and out thru the netserver. > sales rep always told us that digital calls MUST terminate to the > netserver/HyperArc.... not true, although they go thru the netserver/arc as a packet bus/router. > > We have 4 racks which are now populated with quads (analog/digital) and > Network Management Card. Currently they are connected to the POTS network > and to cisco 2511 routers via NIC. We would like to continue to use the > 2511s but would like to get rid of the POTS, connect to PRI and be able to > handle both analog and digital calls. Is that a dream? > Not at all, just plug a PRI in the PRI card and you can handle analog digital cards on the quad modems, assuming you have newer modem code and not much older quad code. > Asking for more: > would it be feasible to have some quads receiving calls from the POTS (NIC) > and others from the PRI card con the same chassis? totally possible. > This would be an even nicer dream for the transition. > > Cheers > > Sergio Manzi > GP Net S.r.l. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Framed-Route, which one?
From: Brian <signal@shreve.net>
Date: 1998-02-12 09:02:51
On Thu, 12 Feb 1998, MegaZone wrote: > Once upon a time Brian shaped the electrons to say... > >signal Authentication-Type = Unix-PW > > Service-Type = Framed, > > Framed-Protocol = PPP, > > Framed-IP-Address = 208.214.45.1, > > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > > Framed-IP-Netmask = 255.255.255.240, > > Framed-MTU = 1500, > > Framed-Compression = Van-Jacobson-TCP-IP > > This looks ok. > > >signal Authentication-Type = Unix-PW > > Service-Type = Framed, > > Framed-Protocol = PPP, > > Framed-IP-Address = 208.214.45.1, > > Framed-Netmask = 255.255.255.255, > > Framed-Route = "208.214.45.0/28 208.214.45.1 1", > > Framed-IP-Netmask = 255.255.255.240, > > Framed-MTU = 1500, > > Framed-Compression = Van-Jacobson-TCP-IP > > This is bogus. Framed-Netmask is the old name for Framed-IP-Netmask, > so you are seting the same thing twice to different values. Thats just it, I meant to type "Framed-IP-Netmask" in that last example above..............in any case, I *thought* you do not declare "Framed-IP-Netmask" when you declare a "Framed-Route".......... Brian > > -MZ > -- > <URL:mailto:megazone@livingston.com> Gweep, author, webmaster, human being, me > <URL:mailto:megazone@megazone.org> H:510-527-0944 W:800-458-9966 510-426-0770 > <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia! > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) DNIS call map to service
From: Douglas Palmer <telos@gain-ny.com>
Date: 1998-02-12 09:17:45
This is probably fairly elementary, but I can't locate the answer in the docs. I want a certain called number to automatically telnet to a certain address. For example, if a user dials into 718-555-4429, I want that user to automatically telnet to 156.121.32.227. No authentication, no other services -- just telnet to that one address. This should be possible with DNIS call mapping, right? Any help would be appreciated. Thanks for your help. DCP
Subject: (usr-tc) Changing radius ports?
From: Phil Freed <phil@freed.com>
Date: 1998-02-12 09:50:18
Is there a way to change the standard Radius authentication and accounting ports in a TC Netserver card? Thanks.... --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."
Subject: (usr-tc) Idle Timeouts revisited
From: Wayne Barber <barberw@tidewater.net>
Date: 1998-02-12 10:18:29
I just had a customer complain about her PC being connected to us overnight. She didn't login, but probably sat at the login: prompt the entire time. We have idle-timeout set in the netserver to 60 minutes, but it didn't drop her. We tested it ourselves just now and we didn't get dropped either. Is there anything special that needs to be set or does this just not work? This is a standard Netserver with 4 mb running 3.3.3 Thanks, Wayne Barber - barberw@tidewater.net Internet System Administrator Coastal Telco Services
Subject: Re: (usr-tc) How do you get PRI ISDN calls to terminate in the modem, rather than , the netserver?
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-12 10:22:07
Thus spake Brian >On Thu, 12 Feb 1998, Sergio Manzi wrote: >> On Wednesday, February 11, 1998 22:54, Pete Ashdown >> [SMTP:pashdown@xmission.com] wrote: >> > Jaye Mathisen said once upon a time: >> > > >> > >How do you get PRI ISDN calls to terminate in the modem, rather than , >> > >the netserver? >> > >> > Set ISDN GW Slot to 0 in the PRI card configuration. >> >> Do you mean that it is feasible to to route ISDN (I mean true ISDN, >> digital, PPP encapsulated calls) to the quad modem and from that to the >> async port on the NIC? We've been asking for that for months and our USR >the nics not involved. You go from the PRI card, to the Modem NAC, and >out thru the netserver. the nic *can* be involved, if you don't want to use a netserver, yes, you can (is my understanding) have the quad modems answer ISDN and send async ppp out the nic card to a seperate access server/terminal server. >> sales rep always told us that digital calls MUST terminate to the >> netserver/HyperArc.... >not true, although they go thru the netserver/arc as a packet bus/router. They *can* go through the netserver/arc as an access server, but they can also go out the nic cards in the back to a seperate box to fulfill the access server/terminal server function. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) routing question
From: Brian <signal@shreve.net>
Date: 1998-02-12 10:28:59
This 255.255.255.0 route is still haunting me, here is the situation: User Configured as: signal Authentication-Type = Unix-PW Service-Type = Framed, Framed-Protocol = PPP, Framed-IP-Address = 208.214.45.1, Framed-IP-Netmask = 255.255.255.255, Framed-Route = "208.214.45.0/28 208.214.45.1 1", Framed-MTU = 1500, Framed-Compression = Van-Jacobson-TCP-IP (With or without the Framed-IP-Netmask this happens) UNIX Server Configured with: $ cat /etc/gated.conf rip yes { nobroadcast ; } ; TC Hub Configured as: set net0 routing on set ripv2 on set enh_routing on yet when I do a netstat -r on the above UNIX server, I get: 208.214.45.0 usr3ts1.shreve. 255.255.255.0 UG eth0 1. The UNIX servers I have are all configured for "nobroadcast", that is to say they don't send out RIP information at all. 2. The TC Hubs are all configured as shown above 3. The only other RIP speaking device on the network, is our Cisco: router rip version 2 network 208.206.76.0 timers basic 30 30 2 60 300 This is racking my brain. On the cisco, I did not add any entries like: ip route 208.214.45.0 255.255.255.240 Ethernet0 ip route 208.214.45.16 255.255.255.240 Ethernet0 ip route 208.214.45.32 255.255.255.240 Ethernet0 . . . I didn't do this, because I just let the Framed-Route entry sent to the TC Hub, add a route to the TC hub, which then broadcasts out to the UNIX Servers (running gated) and the Cisco. Do you think I should explicitly declare routes as shown above on the Cisco, so that the Cisco will send those out regardless of the Framed-Route entry? I would much rather leave the Cisco config alone and have RIP handle it all, but it seems like somethings not liking me using a .0 subnet......... Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NT RAS 4.0 problems..
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-12 10:32:13
Try this - Disable IP header compression on the NT side. krish On Thu, 12 Feb 1998, Brian wrote: > On Thu, 12 Feb 1998, System Administrator wrote: > > > Suddenly, out of the blue, we have numerous NT 4.0 (Server or Workstation) > > customers who cannot connect to our tcs. All of our techs were stumped, > > and it finally landed on my desk. Hell if I could figured it out today, > > after 3 hours w/ one particular customer. Using Quads/HDMs and > > Netservers, in multiple chassis. Customers claim to be able to connect to > > other ISPs with no problems. 95/3.1/Mac users are peachy. > > > > Customer's modem trains and establishes good connection. Syslog seems > > perfectly normal. Customer is handed IP and DNS. LCP negotiation (as > > viewed through debug 0x71) seems fine, although there are many NAK/ACKs in > > the LCP part before settling down. Cannot ping customer. No data comes > > back, traceroute stops at tc. When customer runs 'ipconfig', they see the > > dynamic IP they were given, proper DNS and everything looks normal (for > > NT, heh). > > > > I doubt its related. But when we went to ARC's/HDM's, we had some NT > customers who could not get on. They were setup to use a script for > authentication rather than PAP. Since on our ARC we don't have it > presenting a login prompt (haven't read up on that yet :) ), there script > would fail and so we reconfigured for PAP. > > Also out of the blue, we had like 15 customers get "cannot negotiate > compatible set of protocols" all in the same day! We simply had them > install DUN 1.2, and the latest winsock upgrade and it worked great. > Reinsstalling tcpip, dun etc didnt work, we had to grab these updates. > > Perhaps getting there service packs up to snuff etc, although a pain in > the ass, is the way to go. I would recommend at least having them install > the latest DUN, which does not come with NT. > > Brian > > > > > Any ideas? > > > > Jesse Sipprell > > Senior Systems Engineer > > Evolution Communications, Inc. > > > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) DNIS call map to service
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-12 10:32:26
At 09:17 AM 2/12/98 -0500, Douglas Palmer wrote: >This is probably fairly elementary, but I can't locate the answer in the docs. > >I want a certain called number to automatically telnet to a certain address. > >For example, if a user dials into 718-555-4429, I want that user to >automatically telnet to 156.121.32.227. No authentication, no other >services -- just telnet to that one address. This should be possible with >DNIS call mapping, right? Any help would be appreciated. > >Thanks for your help. >DCP No that is possible with a radius parameters: Service-Type = Login, Login-Service = Rlogin, Login-IP-Host = xxx.xxx.xxx.xxx Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: (usr-tc) routing problem
From: Brian <signal@shreve.net>
Date: 1998-02-12 10:48:06
Ahh, found the "source" for my bad route: earth:/home/signal# /etc/ripquery -r stargate 64 bytes from stargate.shreve.net(208.206.76.1) version 2: 0.0.0.0/0.0.0.0 router 0.0.0.0 metric 1 tag 0000 208.214.44.0/255.255.255.0 router 0.0.0.0 metric 2 tag 0000 208.214.45.0/255.255.255.0 router 0.0.0.0 metric 2 tag 0000 earth:/home/signal# /etc/ripquery -r usr3ts1 184 bytes from usr3ts1.shreve.net(208.206.76.39) version 2: 208.214.45.0/255.255.255.240 router 0.0.0.0 metric 1 tag 0000 You can see its the cisco (stargate) thats giving that bogus info, I will try and find out why..........Like I said, I am not doing anything in the cisco like: ip route 208.214.45.0 255.255.255.240 Ethernet0 but maybe I should be? I was just letting the Cisco add routes on that network as they are announced. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) routing problem
From: Brian <signal@shreve.net>
Date: 1998-02-12 12:13:16
On Thu, 12 Feb 1998, Charles Sprickman wrote: > Where does the Cisco seem to be getting this route? Is there a > ripquery-like command available there? Do you have "ip subnet zero" > enabled on the cisco? > > Charles Yes ip subnet-zero is enabled on the router. The router is learning this route from the Total Control, as they are the only things listed as sending updates. If I dial into "USR #3" than a route for 208.214.45.0 255.255.255.0 is added to all my "gated" machines with a gateway of USR #3. So I did ripquery, and the only thing that comes back with 208.214.45.0 and that /24 netmask, is the cisco. The cisco knows nothing about 208.214.45.0 however, it must have learned it from the TC hub. Somewhere between the TC hub sending out the ripv2 information and the cisco receving the ripv2 info, it all got hosed. My gated machines dont send out any routing information. I have seen 208.214.45.0 added with the PROPER 255.255.255.240 netmask, my theory is that the TC hub beat the cisco in those cases, but most of the time, the cisco wins and it wants to append a /24 netmask to the v2 routing information for that subnet. I wish I could better track this down. > > ~~~~~~~~~ ~~~~~~~~~~~ > Charles Sprickman Internet Channel > INCH System Administration Team (212)243-5200 > spork@inch.com access@inch.com > > On Thu, 12 Feb 1998, Brian wrote: > > > Date: Thu, 12 Feb 1998 10:48:06 -0600 (CST) > > From: Brian <signal@shreve.net> > > Reply-To: usr-tc@lists.xmission.com > > To: USRobotics TC Mailing List <usr-tc@xmission.com> > > Subject: (usr-tc) routing problem > > > > > > Ahh, found the "source" for my bad route: > > > > earth:/home/signal# /etc/ripquery -r stargate > > 64 bytes from stargate.shreve.net(208.206.76.1) version 2: > > 0.0.0.0/0.0.0.0 router 0.0.0.0 metric 1 > > tag > > 0000 > > 208.214.44.0/255.255.255.0 router 0.0.0.0 metric 2 > > tag > > 0000 > > 208.214.45.0/255.255.255.0 router 0.0.0.0 metric 2 > > tag > > 0000 > > earth:/home/signal# /etc/ripquery -r usr3ts1 > > 184 bytes from usr3ts1.shreve.net(208.206.76.39) version 2: > > 208.214.45.0/255.255.255.240 router 0.0.0.0 metric 1 > > tag > > 0000 > > > > > > You can see its the cisco (stargate) thats giving that bogus info, I will > > try and find out why..........Like I said, I am not doing anything in the > > cisco like: > > > > ip route 208.214.45.0 255.255.255.240 Ethernet0 > > > > but maybe I should be? I was just letting the Cisco add routes on that > > network as they are announced. > > > > Brian > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) routing problem
From: Charles Sprickman <spork@inch.com>
Date: 1998-02-12 12:24:48
Where does the Cisco seem to be getting this route? Is there a ripquery-like command available there? Do you have "ip subnet zero" enabled on the cisco? Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Thu, 12 Feb 1998, Brian wrote: > Date: Thu, 12 Feb 1998 10:48:06 -0600 (CST) > From: Brian <signal@shreve.net> > Reply-To: usr-tc@lists.xmission.com > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Subject: (usr-tc) routing problem > > > Ahh, found the "source" for my bad route: > > earth:/home/signal# /etc/ripquery -r stargate > 64 bytes from stargate.shreve.net(208.206.76.1) version 2: > 0.0.0.0/0.0.0.0 router 0.0.0.0 metric 1 > tag > 0000 > 208.214.44.0/255.255.255.0 router 0.0.0.0 metric 2 > tag > 0000 > 208.214.45.0/255.255.255.0 router 0.0.0.0 metric 2 > tag > 0000 > earth:/home/signal# /etc/ripquery -r usr3ts1 > 184 bytes from usr3ts1.shreve.net(208.206.76.39) version 2: > 208.214.45.0/255.255.255.240 router 0.0.0.0 metric 1 > tag > 0000 > > > You can see its the cisco (stargate) thats giving that bogus info, I will > try and find out why..........Like I said, I am not doing anything in the > cisco like: > > ip route 208.214.45.0 255.255.255.240 Ethernet0 > > but maybe I should be? I was just letting the Cisco add routes on that > network as they are announced. > > Brian > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) routing problem
From: Brian <signal@shreve.net>
Date: 1998-02-12 12:24:49
Still looking at this routing problem, I noticed that "Netservers" continue to send RIP v1 information even when in RIP v2 mode. "HiPerARC's" however only send out RIP v2 when you tell them too: Here I query a USR TC configured as: set net0 routing on set ripv2 on set enh_routing on venus:/home/signal# /etc/ripquery -r usr4ts1 44 bytes from usr4ts1.shreve.net(208.206.76.41) version 2: 208.214.45.241/255.255.255.255 router 0.0.0.0 metric 1 208.232.62.128/255.255.255.192 router 0.0.0.0 metric 1 ok no problem there, now lets do a RIP v1 query: venus:/home/signal# /etc/ripquery -r -1 usr4ts1 424 bytes from usr4ts1.shreve.net(208.206.76.41) version 1: 208.232.62.161 metric 1 208.232.62.130 metric 1 208.232.62.131 metric 16 208.232.62.164 metric 1 208.232.62.133 metric 1 208.232.62.165 metric 1 . . . Why did it even respond? Yet on a HiPer ARC: venus:/home/signal# /etc/ripquery -r usr1ts1 24 bytes from usr1ts1.shreve.net(208.206.76.35) version 2: 208.214.44.0/255.255.255.0 router 0.0.0.0 metric 1 and v1: venus:/home/signal# /etc/ripquery -r -1 usr1ts1 venus:/home/signal# So the TC Hub is reporting v1 routing information. And my Cisco should *only* be responding to v2 routing information........are there any known bugs like this in IOS 11.1? Router is as: shrevenet-gw#show ip proto Routing Protocol is "rip" Sending updates every 30 seconds, next due in 23 seconds Invalid after 30 seconds, hold down 2, flushed after 60 Outgoing update filter list for all interfaces is not set Incoming update filter list for all interfaces is not set Redistributing: rip Default version control: send version 2, receive version 2 Interface Send Recv Key-chain Ethernet0 2 2 Serial0.1 2 2 Routing for Networks: 208.206.76.0 Routing Information Sources: Gateway Distance Last Update 208.206.76.31 120 00:00:22 <---- These are all 208.206.76.35 120 00:00:29 USR hubs 208.206.76.39 120 00:00:29 208.206.76.37 120 00:00:00 208.206.76.41 120 00:00:01 Can't think of how this could be getting so screwed up. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) routing problem
From: Brian <signal@shreve.net>
Date: 1998-02-12 12:27:39
On Thu, 12 Feb 1998, Charles Sprickman wrote: > That's odd, I had a similar problem once with setting RIPv2 on the TC. > Once we got to a level 3 tech, it disappeared for no reason... > > I'd suggest going to comp.dcom.sys.cisco and asking about any commands > that can show you where the update is coming from, and why it might be > interpreted as RIPv1. Those folks are *very* helpful, and very > knowledgeable. I even got a call from a concerned Cisco engineer once. I will do just that. > > I'd also fire up a packet-sniffer in case you need evidence for whichever > tech support dept. you might deal with. That helped immensely in my case. > Just sent a dump to usr showing ripv1 packets spewing forth after setting > the ripv2 param... Thanks > > Good Luck, > > Charles > > ~~~~~~~~~ ~~~~~~~~~~~ > Charles Sprickman Internet Channel > INCH System Administration Team (212)243-5200 > spork@inch.com access@inch.com > > On Thu, 12 Feb 1998, Brian wrote: > > > Date: Thu, 12 Feb 1998 12:13:16 -0600 (CST) > > From: Brian <signal@shreve.net> > > To: Charles Sprickman <spork@inch.com> > > Cc: usr-tc@lists.xmission.com > > Subject: Re: (usr-tc) routing problem > > > > On Thu, 12 Feb 1998, Charles Sprickman wrote: > > > > > Where does the Cisco seem to be getting this route? Is there a > > > ripquery-like command available there? Do you have "ip subnet zero" > > > enabled on the cisco? > > > > > > Charles > > > > Yes ip subnet-zero is enabled on the router. The router is learning this > > route from the Total Control, as they are the only things listed as > > sending updates. > > > > If I dial into "USR #3" than a route for 208.214.45.0 255.255.255.0 is > > added to all my "gated" machines with a gateway of USR #3. > > > > So I did ripquery, and the only thing that comes back with 208.214.45.0 > > and that /24 netmask, is the cisco. The cisco knows nothing about > > 208.214.45.0 however, it must have learned it from the TC hub. > > > > Somewhere between the TC hub sending out the ripv2 information and the > > cisco receving the ripv2 info, it all got hosed. My gated machines dont > > send out any routing information. I have seen 208.214.45.0 added with the > > PROPER 255.255.255.240 netmask, my theory is that the TC hub beat the > > cisco in those cases, but most of the time, the cisco wins and it wants to > > append a /24 netmask to the v2 routing information for that subnet. > > > > I wish I could better track this down. > > > > > > > > ~~~~~~~~~ ~~~~~~~~~~~ > > > Charles Sprickman Internet Channel > > > INCH System Administration Team (212)243-5200 > > > spork@inch.com access@inch.com > > > > > > On Thu, 12 Feb 1998, Brian wrote: > > > > > > > Date: Thu, 12 Feb 1998 10:48:06 -0600 (CST) > > > > From: Brian <signal@shreve.net> > > > > Reply-To: usr-tc@lists.xmission.com > > > > To: USRobotics TC Mailing List <usr-tc@xmission.com> > > > > Subject: (usr-tc) routing problem > > > > > > > > > > > > Ahh, found the "source" for my bad route: > > > > > > > > earth:/home/signal# /etc/ripquery -r stargate > > > > 64 bytes from stargate.shreve.net(208.206.76.1) version 2: > > > > 0.0.0.0/0.0.0.0 router 0.0.0.0 metric 1 > > > > tag > > > > 0000 > > > > 208.214.44.0/255.255.255.0 router 0.0.0.0 metric 2 > > > > tag > > > > 0000 > > > > 208.214.45.0/255.255.255.0 router 0.0.0.0 metric 2 > > > > tag > > > > 0000 > > > > earth:/home/signal# /etc/ripquery -r usr3ts1 > > > > 184 bytes from usr3ts1.shreve.net(208.206.76.39) version 2: > > > > 208.214.45.0/255.255.255.240 router 0.0.0.0 metric 1 > > > > tag > > > > 0000 > > > > > > > > > > > > You can see its the cisco (stargate) thats giving that bogus info, I will > > > > try and find out why..........Like I said, I am not doing anything in the > > > > cisco like: > > > > > > > > ip route 208.214.45.0 255.255.255.240 Ethernet0 > > > > > > > > but maybe I should be? I was just letting the Cisco add routes on that > > > > network as they are announced. > > > > > > > > Brian > > > > > > > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > > > > > > > > > - > > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > > with "unsubscribe usr-tc" in the body of the message. > > > > For information on digests or retrieving files and old messages send > > > > "help" to the same address. Do not use quotes in your message. > > > > > > > > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the message. > > > For information on digests or retrieving files and old messages send > > > "help" to the same address. Do not use quotes in your message. > > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) routing problem
From: Charles Sprickman <spork@inch.com>
Date: 1998-02-12 13:21:08
That's odd, I had a similar problem once with setting RIPv2 on the TC. Once we got to a level 3 tech, it disappeared for no reason... I'd suggest going to comp.dcom.sys.cisco and asking about any commands that can show you where the update is coming from, and why it might be interpreted as RIPv1. Those folks are *very* helpful, and very knowledgeable. I even got a call from a concerned Cisco engineer once. I'd also fire up a packet-sniffer in case you need evidence for whichever tech support dept. you might deal with. That helped immensely in my case. Just sent a dump to usr showing ripv1 packets spewing forth after setting the ripv2 param... Good Luck, Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Thu, 12 Feb 1998, Brian wrote: > Date: Thu, 12 Feb 1998 12:13:16 -0600 (CST) > From: Brian <signal@shreve.net> > To: Charles Sprickman <spork@inch.com> > Cc: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) routing problem > > On Thu, 12 Feb 1998, Charles Sprickman wrote: > > > Where does the Cisco seem to be getting this route? Is there a > > ripquery-like command available there? Do you have "ip subnet zero" > > enabled on the cisco? > > > > Charles > > Yes ip subnet-zero is enabled on the router. The router is learning this > route from the Total Control, as they are the only things listed as > sending updates. > > If I dial into "USR #3" than a route for 208.214.45.0 255.255.255.0 is > added to all my "gated" machines with a gateway of USR #3. > > So I did ripquery, and the only thing that comes back with 208.214.45.0 > and that /24 netmask, is the cisco. The cisco knows nothing about > 208.214.45.0 however, it must have learned it from the TC hub. > > Somewhere between the TC hub sending out the ripv2 information and the > cisco receving the ripv2 info, it all got hosed. My gated machines dont > send out any routing information. I have seen 208.214.45.0 added with the > PROPER 255.255.255.240 netmask, my theory is that the TC hub beat the > cisco in those cases, but most of the time, the cisco wins and it wants to > append a /24 netmask to the v2 routing information for that subnet. > > I wish I could better track this down. > > > > > ~~~~~~~~~ ~~~~~~~~~~~ > > Charles Sprickman Internet Channel > > INCH System Administration Team (212)243-5200 > > spork@inch.com access@inch.com > > > > On Thu, 12 Feb 1998, Brian wrote: > > > > > Date: Thu, 12 Feb 1998 10:48:06 -0600 (CST) > > > From: Brian <signal@shreve.net> > > > Reply-To: usr-tc@lists.xmission.com > > > To: USRobotics TC Mailing List <usr-tc@xmission.com> > > > Subject: (usr-tc) routing problem > > > > > > > > > Ahh, found the "source" for my bad route: > > > > > > earth:/home/signal# /etc/ripquery -r stargate > > > 64 bytes from stargate.shreve.net(208.206.76.1) version 2: > > > 0.0.0.0/0.0.0.0 router 0.0.0.0 metric 1 > > > tag > > > 0000 > > > 208.214.44.0/255.255.255.0 router 0.0.0.0 metric 2 > > > tag > > > 0000 > > > 208.214.45.0/255.255.255.0 router 0.0.0.0 metric 2 > > > tag > > > 0000 > > > earth:/home/signal# /etc/ripquery -r usr3ts1 > > > 184 bytes from usr3ts1.shreve.net(208.206.76.39) version 2: > > > 208.214.45.0/255.255.255.240 router 0.0.0.0 metric 1 > > > tag > > > 0000 > > > > > > > > > You can see its the cisco (stargate) thats giving that bogus info, I will > > > try and find out why..........Like I said, I am not doing anything in the > > > cisco like: > > > > > > ip route 208.214.45.0 255.255.255.240 Ethernet0 > > > > > > but maybe I should be? I was just letting the Cisco add routes on that > > > network as they are announced. > > > > > > Brian > > > > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the message. > > > For information on digests or retrieving files and old messages send > > > "help" to the same address. Do not use quotes in your message. > > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > >
Subject: Re: (usr-tc) Telnet logins
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-12 13:33:32
Mark R. Lindsey said once upon a time: > >This RADIUS entry doesn't work for me; the RADIUS server rejects it, >when I login using that name and password: > > mo Password = "mo" > Service-Type = Login, > Login-Service = Telnet, > Login-IP-Host = 204.252.164.18, > Idle-Timeout = 900 > >When I chop this down to > > mo Password = "mo" > Idle-Timeout = 900 > >It logs me in and starts PPP. But I need the telnet thing to work. >What am I doing wrong? My TCs are used exclusively for PPP, now. Does "Login-Service = Rlogin" work? I've never used Telnet for a Login-Service.
Subject: (usr-tc) Telnet logins
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-12 13:46:00
This RADIUS entry doesn't work for me; the RADIUS server rejects it, when I login using that name and password: mo Password = "mo" Service-Type = Login, Login-Service = Telnet, Login-IP-Host = 204.252.164.18, Idle-Timeout = 900 When I chop this down to mo Password = "mo" Idle-Timeout = 900 It logs me in and starts PPP. But I need the telnet thing to work. What am I doing wrong? My TCs are used exclusively for PPP, now. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: Re: (usr-tc) Telnet/rlogin logins
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-12 13:46:46
Mark R. Lindsey said once upon a time: > >: > mo Password = "mo" >: > Service-Type = Login, >: > Login-Service = Telnet, >: > Login-IP-Host = 204.252.164.18, >: > Idle-Timeout = 900 > >: Does "Login-Service = Rlogin" work? I've never used Telnet for a >: Login-Service. > >No: rlogin doesn't work either. Try doing this command on the Netserver in question first: set all login network dialin
Subject: (usr-tc) Hiper answers differently from Quad
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-12 13:55:20
My HiPer DSP's answer tones are different from my Quads. The HiPer's start with a continous clear tone, much like a 212-A answer tone, then goes into negotiation. My Quads start with a much more screechy tone and seem to negotiate high-speeds faster. The Couriers I have in use sound the same as the Quads in answering. My settings aren't too far off the defaults. Does anyone know what controls this answer method? Which is better?
Subject: Re: (usr-tc) Idle Timeouts revisited
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-12 15:16:12
On Thu, 12 Feb 1998, Wayne Barber wrote: > I just had a customer complain about her PC being connected to us > overnight. She didn't login, but probably sat at the login: prompt > the entire time. We have idle-timeout set in the netserver to 60 > minutes, but it didn't drop her. We tested it ourselves just now and > we didn't get dropped either. Is there anything special that needs to > be set or does this just not work? > > This is a standard Netserver with 4 mb running 3.3.3 Is this a Netserver/8 or /16, or is it a Netserver card in a Total Control rack? If it's a Netserver card, I'd recommend an upgrade to the latest code, yours is pretty old. My Netserver card with the latest code does drop logins after a bit. The newer code requires more memory, so you'll need to put a 16MB 4x32 72 pin (non-parity/non-EDO) SIMM in the Netserver. If it's a Netserver/8 or /16 the above doesn't apply. I don't how you'd fix this problem on a Netserver/8 or /16. Brian
Subject: Re: (usr-tc) Telnet/rlogin logins
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-12 15:37:02
: > mo Password = "mo" : > Service-Type = Login, : > Login-Service = Telnet, : > Login-IP-Host = 204.252.164.18, : > Idle-Timeout = 900 : Does "Login-Service = Rlogin" work? I've never used Telnet for a : Login-Service. No: rlogin doesn't work either.
Subject: Re: (usr-tc) ITU Standard
From: jason_kelton@3com.com
Date: 1998-02-12 15:40:16
Yes... This is Free Free, not Just Free as per the previous upgrade... Or atleast, thats what I've been told to-date. Regards, Jason... ogura@telcel.net.ve on 12/02/98 08:29:36 Please respond to usr-tc@lists.xmission.com cc: phenkle@usr.com (bcc: Jason Kelton/AU/3Com) It?s really free or USR will charge 60$ for upgrade. Remenber V.34 to X2 "Upgradeble"!!! Brian Elfert wrote: > On Wed, 11 Feb 1998 phenkle@usr.com wrote: > > > All x2 customers (Sportsters, Netservers, Total Control, etc.) are > > entitled to a free v.90 upgrade. (after all it is just software) > > > > However, Total Control will auto-negotiate x2 or v.90 calls, so the > > client modems do not have to change :) > > Is the Total Control modem code going to be out before the > Sportster/Courier modem code this time? > > When x2 was released, the client code was available a week or two in > advance, and the ISPs like us, who signed up in advance on the x2 web > site, were pummeled with users asking when we were going to have x2 > available since their modems supported it. > > Will client modems with V.90 be able to drop back to x2 if the server > modem supports only x2 and not V.90? > > Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. begin: vcard fn: Ken Ogura n: Ogura;Ken org: Telcel adr: Av. Francisco de Miranda, Parque Canaima;;Piso 16, Los Palos Grandes;Caracas;Miranda;;Venezuela email;internet: ogura@telcel.net.ve title: Gerente Operaciones T-Net tel;work: 58-2-2018170 tel;fax: 58-2-2018140 tel;home: 58-14-149922 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard
Subject: Re: (usr-tc) Hiper answers differently from Quad
From: Brian <signal@shreve.net>
Date: 1998-02-12 15:40:31
On Thu, 12 Feb 1998, Pete Ashdown wrote: > My HiPer DSP's answer tones are different from my Quads. The HiPer's start > with a continous clear tone, much like a 212-A answer tone, then goes into > negotiation. My Quads start with a much more screechy tone and seem to > negotiate high-speeds faster. The Couriers I have in use sound the same as > the Quads in answering. My settings aren't too far off the defaults. Does > anyone know what controls this answer method? Which is better? > Pete, We noticed this too, and personally I like the nice clear sound of a HDM over a Quad. I wonder what the difference is there, and what the logic was for a different transmit carrier. Brian > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) show netports command
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-12 16:29:57
Can anyone expound on the show netports command and what information it displays? It seems to me, you have to be on the netserver that has the first channel of the bundle to get any information out of it at all...and the information given is: port the "virtual" port, or tunnel port number that the other connections are coming in on. host the "host" of the bundle, which means the netserver that has the first channel of the bundle connected...which since the only way to get any information out of netports, its going to be the netserver that you're connected to and getting the information from...seems that this is a pointless field to have then type shows network since its a network tunnel for MPIP status establish...pretty brainless here...its actually up and running (other status'es may be there that I haven't seen) input and output how much data has flowed over the tunnel connection? pend how much data is queue'd up waiting to go through the tunnel connection? Can anyone confirm or correct this information? Can anyone point out to me what the point of having the "host" show the first link of the MP bundle (always the system you're connecting to), rather than the (at least IMHO) more useful information of the host the this channel is connecting *from*? -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) NT RAS 4.0 problems..
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-12 20:36:35
Suddenly, out of the blue, we have numerous NT 4.0 (Server or Workstation) customers who cannot connect to our tcs. All of our techs were stumped, and it finally landed on my desk. Hell if I could figured it out today, after 3 hours w/ one particular customer. Using Quads/HDMs and Netservers, in multiple chassis. Customers claim to be able to connect to other ISPs with no problems. 95/3.1/Mac users are peachy. Customer's modem trains and establishes good connection. Syslog seems perfectly normal. Customer is handed IP and DNS. LCP negotiation (as viewed through debug 0x71) seems fine, although there are many NAK/ACKs in the LCP part before settling down. Cannot ping customer. No data comes back, traceroute stops at tc. When customer runs 'ipconfig', they see the dynamic IP they were given, proper DNS and everything looks normal (for NT, heh). Any ideas? Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Idle Timeouts revisited
From: Charles Hill <chill@ionet.net>
Date: 1998-02-12 20:54:35
Set the idle timeouts in the modems. . . if the netserver doesn't timeout, the modem will. -CH On Thu, 12 Feb 1998, Wayne Barber wrote: > I just had a customer complain about her PC being connected to us > overnight. She didn't login, but probably sat at the login: prompt > the entire time. We have idle-timeout set in the netserver to 60 > minutes, but it didn't drop her. We tested it ourselves just now and > we didn't get dropped either. Is there anything special that needs to > be set or does this just not work? > > This is a standard Netserver with 4 mb running 3.3.3 > > Thanks, > > > Wayne Barber - barberw@tidewater.net > Internet System Administrator > Coastal Telco Services > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) NT RAS 4.0 problems..
From: Brian <signal@shreve.net>
Date: 1998-02-12 21:05:42
On Thu, 12 Feb 1998, System Administrator wrote: > Suddenly, out of the blue, we have numerous NT 4.0 (Server or Workstation) > customers who cannot connect to our tcs. All of our techs were stumped, > and it finally landed on my desk. Hell if I could figured it out today, > after 3 hours w/ one particular customer. Using Quads/HDMs and > Netservers, in multiple chassis. Customers claim to be able to connect to > other ISPs with no problems. 95/3.1/Mac users are peachy. > > Customer's modem trains and establishes good connection. Syslog seems > perfectly normal. Customer is handed IP and DNS. LCP negotiation (as > viewed through debug 0x71) seems fine, although there are many NAK/ACKs in > the LCP part before settling down. Cannot ping customer. No data comes > back, traceroute stops at tc. When customer runs 'ipconfig', they see the > dynamic IP they were given, proper DNS and everything looks normal (for > NT, heh). > I doubt its related. But when we went to ARC's/HDM's, we had some NT customers who could not get on. They were setup to use a script for authentication rather than PAP. Since on our ARC we don't have it presenting a login prompt (haven't read up on that yet :) ), there script would fail and so we reconfigured for PAP. Also out of the blue, we had like 15 customers get "cannot negotiate compatible set of protocols" all in the same day! We simply had them install DUN 1.2, and the latest winsock upgrade and it worked great. Reinsstalling tcpip, dun etc didnt work, we had to grab these updates. Perhaps getting there service packs up to snuff etc, although a pain in the ass, is the way to go. I would recommend at least having them install the latest DUN, which does not come with NT. Brian > Any ideas? > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Telnet logins
From: Charles Hill <chill@ionet.net>
Date: 1998-02-12 21:14:55
On Thu, 12 Feb 1998, Mark R. Lindsey wrote: > This RADIUS entry doesn't work for me; the RADIUS server rejects it, > when I login using that name and password: > > mo Password = "mo" > Service-Type = Login, > Login-Service = Telnet, > Login-IP-Host = 204.252.164.18, > Idle-Timeout = 900 Mark, This one works for me: uucp Password = "UNIX" User-Service-Type = Login-User, Login-Service = Telnet, Login-Host = xxx.xxx.xxx.xxx -CH
Subject: Re: (usr-tc) NT RAS 4.0 problems..
From: William Behrens <wbehrens@feist.com>
Date: 1998-02-12 21:45:27
Is the customer using TCP header compression ? We have found that by turning off compression on the NT client that this problem is cured. I don't know why this is (other than NT RAS maybe more configurable at a lower level than Win95) but it seems to fix 98% of my NT user problems with exactly what you've described. William Behrens Director of Network Operations ParaCom Technologies Inc. www.paracom.com
Subject: Re: (usr-tc) Changing radius ports?
From: Phil Freed <phil@freed.com>
Date: 1998-02-13 04:06:23
I wrote: > Is there a way to change the standard Radius authentication and > accounting ports in a TC Netserver card? Yes, of course there is, and (naturally enough) I found it in the docs shortly after posting this question. FWIW, the answer is set auth host/port set account host/port --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."
Subject: (usr-tc) Performance on Chan T1 vs PRI vs ISDN
From: Phil Freed <phil@freed.com>
Date: 1998-02-13 04:06:24
We have found that users with 33.6 modems cannot achieve 33.6 when they connect through a channelized T1 using AMI (as opposed to B8ZS). We've been told that 33.6 connections are impossible under these conditions because of the bandwidth lost to Robbed Bit Signalling. But we've also been told by folks with PM3s (well, at least on such "folk") that they regularly get 33.6 connections under these circumstances. (Of course, connection speed isn't the most important thing, since it can always be renegotiated. But it _is_ what most of our users see, and it's what we get lots of support calls on.) I was wondering what experience others have using various kinds of lines - especially channelized T1s - on TCs or other NASs. I'd be interested in hearing about both connection speeds and negotiated speeds. In fact, if folks just want to dump stats, then send them directly to me: pfreed@cybertours.com and I'll summarize for the list in a week or so. While I'm on the subject: I've been told that you will seldom (if ever) see the upstream portion of an X2 connection go over 28.8. Is this true? Anyone have any thoughts on why it might be so? Thanks for your time. --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."
Subject: Re: (usr-tc) Telnet logins
From: MegaZone <megazone@megazone.org>
Date: 1998-02-13 04:22:09
Once upon a time Mark R. Lindsey shaped the electrons to say... >This RADIUS entry doesn't work for me; the RADIUS server rejects it, >when I login using that name and password: > > mo Password = "mo" > Service-Type = Login, > Login-Service = Telnet, > Login-IP-Host = 204.252.164.18, > Idle-Timeout = 900 I bet you have the ports on the TC setup for network logins but not shell logins. You need to have the ports support shell logins. -MZ -- <URL:mailto:megazone@livingston.com> Gweep, author, webmaster, human being, me <URL:mailto:megazone@megazone.org> H:510-527-0944 W:800-458-9966 510-426-0770 <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia!
Subject: Re: (usr-tc) Telnet logins
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-13 08:08:29
: > mo Password = "mo" : > Service-Type = Login, : > Login-Service = Telnet, : > Login-IP-Host = 204.252.164.18, : > Idle-Timeout = 900 : : I bet you have the ports on the TC setup for network logins but not : shell logins. You need to have the ports support shell logins. Hrm. Well, yeah, as a general rule, I do something like set all login network dialin which has seemed to work. Is it possible to configure the port for both PPP login and Service-Type's besides PPP? (I know it's been done. I really want to know how.) Thanks, all. Mark
Subject: Re: (usr-tc) NT RAS 4.0 problems..
From: chris loelke <chris@okanagan.net>
Date: 1998-02-13 09:38:39
we had the same problem, turn MS compression off at the workstation At 08:36 PM 2/12/98 -0500, you wrote: >Suddenly, out of the blue, we have numerous NT 4.0 (Server or Workstation) >customers who cannot connect to our tcs. All of our techs were stumped, >and it finally landed on my desk. Hell if I could figured it out today, >after 3 hours w/ one particular customer. Using Quads/HDMs and >Netservers, in multiple chassis. Customers claim to be able to connect to >other ISPs with no problems. 95/3.1/Mac users are peachy. > >Customer's modem trains and establishes good connection. Syslog seems >perfectly normal. Customer is handed IP and DNS. LCP negotiation (as >viewed through debug 0x71) seems fine, although there are many NAK/ACKs in >the LCP part before settling down. Cannot ping customer. No data comes >back, traceroute stops at tc. When customer runs 'ipconfig', they see the >dynamic IP they were given, proper DNS and everything looks normal (for >NT, heh). > >Any ideas? > >Jesse Sipprell >Senior Systems Engineer >Evolution Communications, Inc. > >* Finger sysadmin@evcom.net for my PGP Public Key * > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) How do you get PRI ISDN calls to terminate in the
From: jason_kelton@3com.com
Date: 1998-02-13 10:01:39
To elaborate, You need to tell the Quad Modems to talk to the ISDN PRI card. for T1/E1 R2, type AT%D1 and for T1/E1 PRI type AT%D2 on each modem port. Alternatively, via the NMC using TCM, you can change this in the Line Interface Source, in the Line interface settings to your specific choice. For termination of both analog and digital calls, you must be running TCS 2.5.1 components or later. Also in the PRI config, do not set a default ISDN Gateway slot, and make sure allow analog modem calls is enabled. What should happen is when a call comes in, the PRI identifies from the D-Channel, what type of call is coming in, and routes the call to the first available device. If no devices are available, then an NU (busy) signal will be sent back to the customer. You can do the "Dream" as mentioned. Those that wish to continue to use the POTS, leave them as is, and just apply the above settings to those you want terminated on the PRI. Regards, Jason Kelton Carrier Systems Engineer 3Com ANZA. Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) modem, rather than , the netserver? Thus spake Brian >On Thu, 12 Feb 1998, Sergio Manzi wrote: >> On Wednesday, February 11, 1998 22:54, Pete Ashdown >> [SMTP:pashdown@xmission.com] wrote: >> > Jaye Mathisen said once upon a time: >> > > >> > >How do you get PRI ISDN calls to terminate in the modem, rather than , >> > >the netserver? >> > >> > Set ISDN GW Slot to 0 in the PRI card configuration. >> >> Do you mean that it is feasible to to route ISDN (I mean true ISDN, >> digital, PPP encapsulated calls) to the quad modem and from that to the >> async port on the NIC? We've been asking for that for months and our USR >the nics not involved. You go from the PRI card, to the Modem NAC, and >out thru the netserver. the nic *can* be involved, if you don't want to use a netserver, yes, you can (is my understanding) have the quad modems answer ISDN and send async ppp out the nic card to a seperate access server/terminal server. >> sales rep always told us that digital calls MUST terminate to the >> netserver/HyperArc.... >not true, although they go thru the netserver/arc as a packet bus/router. They *can* go through the netserver/arc as an access server, but they can also go out the nic cards in the back to a seperate box to fulfill the access server/terminal server function. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456 - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) Radius crashes
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-13 10:40:48
It does cause it to reboot. There was a patch also to fix it - It is one line change in the radiusd.c file. krish On Fri, 13 Feb 1998, MegaZone wrote: > Once upon a time Mike shaped the electrons to say... > >I think this is a bug in Livingston radius.. When a NS reboots it sends a > > It is odd - since an Accounting-On packet doesn't make us reboot. It is > just logged as: > Acct-Status-Type = 7 > > It should not cause a reboot, and I've seen this be logged and there are > many users running TCs against Lucent RADIUS without reboots. > > I would run 'radiusd -x' and then do the reboot - see what RADIUS is > receiving. > > -MZ > -- > <URL:mailto:megazone@megazone.org> Gweep, author, webmaster, human being, me > "A little nonsense now and then, is relished by the wisest men" 510-527-0944 > <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia! > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) 3.7.24 problem
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-13 12:05:37
On Fri, 13 Feb 1998, Pete Ashdown wrote: > Jeff Mcadams said once upon a time: > > >I just want some decent quality control out of USR on these code > >releases! How hard is it to check whether a zmodem file transfer > >works!? Zmodem does work with Telnet, I am using a linux machine and I use the command sz -e to start the download. I don't see the problem - What host are you using? Have you tried with escape code for your zmodem version? krish > > As a friend in the software industry put it, it doesn't meet the test > criteria. Since the majority of MIS managers out there are using the same > code their TC shipped with and don't know any different, the robustness of > TC code doesn't get tested. It is fair assumption that just about everyone > subscribed to this list squeezes their equipment to the limit and finds > bugs that most people never encounter. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) Idle Timeouts revisited
From: jason_kelton@3com.com
Date: 1998-02-13 12:43:51
Brian, I think the problem here, is that the user is using the old Standard NETServers without the FR interfaces and Munich Daughterboard. If thats the case, then 3.3.3 is the latest!! :) Everything got migrated over to the NETServer PRI's which now require 16MB to operate... Regards, Jason Kelton. brian@citilink.com on 13/02/98 07:16:12 Please respond to usr-tc@lists.xmission.com cc: usr-tc@lists.xmission.com (bcc: Jason Kelton/AU/3Com) On Thu, 12 Feb 1998, Wayne Barber wrote: > I just had a customer complain about her PC being connected to us > overnight. She didn't login, but probably sat at the login: prompt > the entire time. We have idle-timeout set in the netserver to 60 > minutes, but it didn't drop her. We tested it ourselves just now and > we didn't get dropped either. Is there anything special that needs to > be set or does this just not work? > > This is a standard Netserver with 4 mb running 3.3.3 Is this a Netserver/8 or /16, or is it a Netserver card in a Total Control rack? If it's a Netserver card, I'd recommend an upgrade to the latest code, yours is pretty old. My Netserver card with the latest code does drop logins after a bit. The newer code requires more memory, so you'll need to put a 16MB 4x32 72 pin (non-parity/non-EDO) SIMM in the Netserver. If it's a Netserver/8 or /16 the above doesn't apply. I don't how you'd fix this problem on a Netserver/8 or /16. Brian - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) Idle Timeouts revisited
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-13 12:59:29
Charles Hill said once upon a time: > > >Set the idle timeouts in the modems. . . if the netserver doesn't timeout, >the modem will. -CH Unfortunately, this is a lousy solution for situations where you need to have no timeouts and timeouts mixed together.
Subject: Re: (usr-tc) Changing radius ports?
From: jason_kelton@3com.com
Date: 1998-02-13 13:05:28
Yep... use the following command line SET AUTHENTIC <IpAddr> <port> <secret> <Radius Ver> I think a forward slash is required between the IP and the port 123.456.789.xxx/1600 I believe its documented in the NETServer manual Regards, Jason. phil@freed.com on 14/02/98 01:50:18 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) Is there a way to change the standard Radius authentication and accounting ports in a TC Netserver card? Thanks.... --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific." - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) 3.7.24 problem
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-13 14:02:21
Jeff Mcadams said once upon a time: >This royally sucks, as we have customers that are still using PPP on our >shell systems that depend on an 8-bit clean connection, and we *really* >don't want to use rlogin because of the hideous, horrible security holes >which that opens up... > >Any thoughts? What holes are you talking about? I've been using rlogin for dialin with tcp wrapper restrictions since 1993. No holes that I've seen.
Subject: Re: (usr-tc) 3.7.24 problem
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-13 14:49:18
Jeff Mcadams said once upon a time: >>What holes are you talking about? I've been using rlogin for dialin with >>tcp wrapper restrictions since 1993. No holes that I've seen. > >The r commands are basically just inherently insecure. :) UNIX is inherently insecure. You just have to administrate it properly. :-) >tcp wrappers mitigates the problem considerably, but the r command setup >is just inherently insecure (we've a tad bit fascist around here)...ie, >the ability to login without any sort of authentication (possibly), or >at the best host based authentication. Depending on what is in your hosts.equiv and .rhosts. We run regular searches against this type of thing for our users, and rlogin isn't possible for any of the administrative accounts. >Also, and this is incidental given what I suspect your setup is...there >was a buffer overrun problem in the actual rlogin client command on >Solaris fairly recently (check back issues of bugtraq for more info), >and since its setuid root....well...'nuff said. :) Sun issued a patch for this quite a while ago. I also don't use the stock Solaris rlogind, but a modified version of the one in Log-Daemon (back to the source issue). You've also got to recall that "ping" had the same problem. >The other issue we have with possibly using rlogin, is that it would >break our user's dial-in scripts. :/ Thats a bigger issue. Personally though, I have problems with the cleanliness of telnet, which is why I don't use it for terminal sessions.
Subject: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-13 15:19:41
AIGH!!!!!!!! OK, 3.7.24 looks like it *might* have fixed the problem we've had with customers using MLPPP with MPIP to bond dialing in, *BUT*... it appears that this version of code no longer has an 8-bit clean telnet! This royally sucks, as we have customers that are still using PPP on our shell systems that depend on an 8-bit clean connection, and we *really* don't want to use rlogin because of the hideous, horrible security holes which that opens up... Any thoughts? -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) security
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-13 15:34:18
System Administrator said once upon a time: >2. Filter your own IPs at your border routers so nobody spoofs 'em coming >in from the outside world. Filter 127.0.0.0/8 as well. Prevents a great >deal of blatant NFS misconfiguration security holes. :) Just wanted to add something to this while we're on the subject. There are several other "internal only" subnets you should filter as well. Here's a snippet of my incoming Cisco filter: access-list 112 deny ip host 0.0.0.0 any access-list 112 deny ip host 255.255.255.255 any access-list 112 deny ip 127.0.0.0 0.255.255.255 any access-list 112 deny ip 10.0.0.0 0.255.255.255 any access-list 112 deny ip 172.16.0.0 0.15.255.255 any access-list 112 deny ip 192.0.2.0 0.0.0.255 any access-list 112 deny ip 192.168.0.0 0.0.255.255 any You would be _amazed_ at how often these are hit, especially the loopback address. The other thing that is important is to put "no ip directed-broadcast" on all your internal LAN interfaces!
Subject: Re: (usr-tc) 3.7.24 problem
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-13 15:45:37
Jeff Mcadams said once upon a time: >I just want some decent quality control out of USR on these code >releases! How hard is it to check whether a zmodem file transfer >works!? As a friend in the software industry put it, it doesn't meet the test criteria. Since the majority of MIS managers out there are using the same code their TC shipped with and don't know any different, the robustness of TC code doesn't get tested. It is fair assumption that just about everyone subscribed to this list squeezes their equipment to the limit and finds bugs that most people never encounter.
Subject: (usr-tc) Radius crashes
From: James Wilson <james@cruxnet.com>
Date: 1998-02-13 16:24:50
It seems that when we restart a tc hub, radius crashes and needs to be restarted before anyone can login. Has anyone else seen this? We are using Livingston Radius 2.0.1. James
Subject: Re: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-13 16:28:48
Thus spake Pete Ashdown >Jeff Mcadams said once upon a time: >>This royally sucks, as we have customers that are still using PPP on our >>shell systems that depend on an 8-bit clean connection, and we *really* >>don't want to use rlogin because of the hideous, horrible security holes >>which that opens up... >>Any thoughts? >What holes are you talking about? I've been using rlogin for dialin with >tcp wrapper restrictions since 1993. No holes that I've seen. The r commands are basically just inherently insecure. :) tcp wrappers mitigates the problem considerably, but the r command setup is just inherently insecure (we've a tad bit fascist around here)...ie, the ability to login without any sort of authentication (possibly), or at the best host based authentication. Also, and this is incidental given what I suspect your setup is...there was a buffer overrun problem in the actual rlogin client command on Solaris fairly recently (check back issues of bugtraq for more info), and since its setuid root....well...'nuff said. :) The other issue we have with possibly using rlogin, is that it would break our user's dial-in scripts. :/ -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Radius crashes
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-13 16:28:58
At 04:24 PM 2/13/98 -0400, you wrote: >It seems that when we restart a tc hub, radius crashes and needs to be >restarted before anyone can login. Has anyone else seen this? We are using >Livingston Radius 2.0.1. > I think this is a bug in Livingston radius.. When a NS reboots it sends a RADIUS accounting packet with the Acct-Status-Type attribute set to 7. No RADIUS packet should cause a server to die, it should just throw it out if it doesn't understand the packet. (From the RFC) Type 40 for Acct-Status-Type Length 6 Value The value field is four octets 1 Start 2 Stop 7 Accounting-On 8 Accounting-Off -m
Subject: Re: (usr-tc) 3.7.24 problem
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-13 16:41:23
: This royally sucks, as we have customers that are still using PPP on our : shell systems that depend on an 8-bit clean connection, and we *really* : don't want to use rlogin because of the hideous, horrible security holes : which that opens up... Use TCP wrappers on the Unix system to restrict connections to the rlogin service to your netservers.
Subject: Re: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-13 16:46:23
Thus spake Mark R. Lindsey >: This royally sucks, as we have customers that are still using PPP on our >: shell systems that depend on an 8-bit clean connection, and we *really* >: don't want to use rlogin because of the hideous, horrible security holes >: which that opens up... >Use TCP wrappers on the Unix system to restrict connections to the rlogin >service to your netservers. Well, that mitigates the security issue (for the most part anyway), but it still breaks customer scripts, so, no go. I just want some decent quality control out of USR on these code releases! How hard is it to check whether a zmodem file transfer works!? -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-13 17:02:13
Thus spake Pete Ashdown >Jeff Mcadams said once upon a time: >>>What holes are you talking about? I've been using rlogin for dialin with >>>tcp wrapper restrictions since 1993. No holes that I've seen. >>The r commands are basically just inherently insecure. :) >UNIX is inherently insecure. You just have to administrate it >properly. :-) Indeed, and part of administering it properly is not supporting inherently insecure protocols. ;) >>tcp wrappers mitigates the problem considerably, but the r command setup >>is just inherently insecure (we've a tad bit fascist around here)...ie, >>the ability to login without any sort of authentication (possibly), or >>at the best host based authentication. >Depending on what is in your hosts.equiv and .rhosts. We run regular >searches against this type of thing for our users, and rlogin isn't >possible for any of the administrative accounts. Yes, that does help security again, but how hard is it to use a linux box and spoof a system name and get access to an account that shouldn't have access to? There are innumerable scenarios for the rcommands because they are inherently insecure. We use ssh in place of the rcommands here... *shrug* tis a choice that each site must use. >>Also, and this is incidental given what I suspect your setup is...there >>was a buffer overrun problem in the actual rlogin client command on >>Solaris fairly recently (check back issues of bugtraq for more info), >>and since its setuid root....well...'nuff said. :) >Sun issued a patch for this quite a while ago. I also don't use the stock >Solaris rlogind, but a modified version of the one in Log-Daemon (back to >the source issue). You've also got to recall that "ping" had the same >problem. Yup, and we had already replaced ping on our systems...as well as ps and others that had buffer overflow problems...had a grand total of about 15 setuid programs on our system. >>The other issue we have with possibly using rlogin, is that it would >>break our user's dial-in scripts. :/ >Thats a bigger issue. Personally though, I have problems with the >cleanliness of telnet, which is why I don't use it for terminal sessions. *shrug* to each his own...its actually cleaner here based on our setup, though we've never had what I would call a truly clean setup with USR equip. (or with Livingston, or Xyplex, or...or...or...) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) 3.7.24 problem
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-13 17:12:30
On Fri, 13 Feb 1998, Pete Ashdown wrote: > >tcp wrappers mitigates the problem considerably, but the r command setup > >is just inherently insecure (we've a tad bit fascist around here)...ie, > >the ability to login without any sort of authentication (possibly), or > >at the best host based authentication. > > Depending on what is in your hosts.equiv and .rhosts. We run regular > searches against this type of thing for our users, and rlogin isn't > possible for any of the administrative accounts. Right. Three steps to making rlogins secure: 1. TCP wrappers -- rlogin should only be needed from your tcrs/trusted-*local* hosts. Others can use ssh, it's much better anyway. 2. Filter your own IPs at your border routers so nobody spoofs 'em coming in from the outside world. Filter 127.0.0.0/8 as well. Prevents a great deal of blatant NFS misconfiguration security holes. :) 2b. (not related to rlogin security, at least not at *your site*). Filter your _outgoing_ IPs so that your customer's can't use somebody else's IP. Hell, if everyone did this, IP spoofing wouldn't even be an issue. 3. Most rloginds can be passed an arguement to completely ignore .rhosts in user directories. Good idea. Check your man page and alter your inetd.conf/xinetd.conf to suit. hosts.equiv is still available, but that shouldn't be world writable (or readable) anyway. That's it. If you do the above, rlogind isn't really a security issue anymore (although the actual setuid rlogin utility [used for rlogins outbound] is a different matter, depending on your platform). It's not even necessary for rlogin to be on your system in order for tcrs to rlogin to your unix boxen, you just need rlogind. PS. I'd like to plug xinetd for a second. :) It's really nifty, smaller footprint than inetd *and* built-in tcp wrappers. Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-13 17:48:27
Thus spake Pete Ashdown >Jeff Mcadams said once upon a time: >>I just want some decent quality control out of USR on these code >>releases! How hard is it to check whether a zmodem file transfer >>works!? >As a friend in the software industry put it, it doesn't meet the test >criteria. Since the majority of MIS managers out there are using the same >code their TC shipped with and don't know any different, the robustness of >TC code doesn't get tested. It is fair assumption that just about everyone >subscribed to this list squeezes their equipment to the limit and finds >bugs that most people never encounter. You'd think they'd pay more attention to us then wouldn't you? -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Radius crashes (fwd)
From: Jim Riffle <rif@rif.hoosierlink.net>
Date: 1998-02-13 18:11:16
> I think this is a bug in Livingston radius.. When a NS reboots it sends a > RADIUS accounting packet with the Acct-Status-Type attribute set to 7. No > RADIUS packet should cause a server to die, it should just throw it out if > it doesn't understand the packet. The NS appears to send that to all the radius servers listed (primary, secondary, accounting). Is it necessary for this to be sent to the primary and secondary auth servers if they differ from the accounting server? Jim --- Jim Riffle rif@hoosierlink.net HoosierLink & KC Online Network/Systems Administrator
Subject: Re: (usr-tc) Radius crashes
From: MegaZone <megazone@megazone.org>
Date: 1998-02-13 19:38:51
Once upon a time Mike shaped the electrons to say... >I think this is a bug in Livingston radius.. When a NS reboots it sends a It is odd - since an Accounting-On packet doesn't make us reboot. It is just logged as: Acct-Status-Type = 7 It should not cause a reboot, and I've seen this be logged and there are many users running TCs against Lucent RADIUS without reboots. I would run 'radiusd -x' and then do the reboot - see what RADIUS is receiving. -MZ -- <URL:mailto:megazone@megazone.org> Gweep, author, webmaster, human being, me "A little nonsense now and then, is relished by the wisest men" 510-527-0944 <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia!
Subject: Re: (usr-tc) 3.7.24 problem
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-13 20:34:52
On Sat, 14 Feb 1998, Jeff Mcadams wrote: > Thus spake Tatai SV Krishnan > > >On Fri, 13 Feb 1998, Pete Ashdown wrote: > >> Jeff Mcadams said once upon a time: > >> > >> >I just want some decent quality control out of USR on these code > >> >releases! How hard is it to check whether a zmodem file transfer > >> >works!? > > >Zmodem does work with Telnet, I am using a linux machine and I use the > >command sz -e to start the download. I don't see the problem - What host > >are you using? Have you tried with escape code for your zmodem version? > > If telnet were 8-bit clean (binary mode) as you advertise, I shouldn't > *have* to use -e, or at least I've never had to in the past! > Regardless, running ppp and uucp over telnet also broke with this upgrade > (though as I mentioned in another post, we do have a workaround now) > indicating a lack of a binary connection as well. Telnet is 7 bit, binary telnet is 8-bit. We support both. You are not considering the problems on your host side, if you host side support binary telnet you need not do -e. What I suggest is not a workarond for our box, it is a workaround for the host. > > Again, we're running some (old) accounts with ppp on our shell servers, > and also have uucp accounts on the shell servers, and were using telnet > (previously binary mode) from the netservers to get to them, we've had > to come up with another way to do it since 3.7.24 broke the binary mode > support in telnet. Atleast from what I have seen so far with 3.7.24 code I have not seen anything wrong with binary telnet. I will however check it out on monday and let you know ( just need to test a litte more :-)... ) If its broken we will fix it. krish > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) 3.7.24 problem
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-13 21:31:42
On Sat, 14 Feb 1998, Jeff Mcadams wrote: > Thus spake Tatai SV Krishnan > >On Sat, 14 Feb 1998, Jeff Mcadams wrote: > >> If telnet were 8-bit clean (binary mode) as you advertise, I shouldn't > >> *have* to use -e, or at least I've never had to in the past! > >> Regardless, running ppp and uucp over telnet also broke with this upgrade > >> (though as I mentioned in another post, we do have a workaround now) > >> indicating a lack of a binary connection as well. > > >Telnet is 7 bit, binary telnet is 8-bit. We support both. > > Not in 3.7.24 you don't. :) That's to be seen. I will investigate on Monday, as of now I do not see any problems with my host and 3.7.24 code on my NETServer. Will certainly investigate. > > >You are not > >considering the problems on your host side, if you host side support > >binary telnet you need not do -e. > > Have been able to do this from our host with all the terminal server > equipment we've ever run (Xyplex, Livingston, and USR before 3.7.24), > when we put 3.7.24 on, sz without the -e quit working...as well as ppp > and uucp from hosts over telnet. > > >What I suggest is not a workarond for our box, it is a workaround for the > >host. > > We tried that since we had source code to telnetd, but couldn't change > anything that would fix it...our conclusion was that the telnet on the > netserver was broken WRT 8-bit'edness. > Thats fair, but still - I will have to test. > >> Again, we're running some (old) accounts with ppp on our shell servers, > >> and also have uucp accounts on the shell servers, and were using telnet > >> (previously binary mode) from the netservers to get to them, we've had > >> to come up with another way to do it since 3.7.24 broke the binary mode > >> support in telnet. > > >Atleast from what I have seen so far with 3.7.24 code I have not seen > >anything wrong with binary telnet. I will however check it out on > >monday and let you know ( just need to test a litte more :-)... ) > >If its broken we will fix it. > > Yup, you need to look farther then, 'cause its definitely broken. :) > Will do - You are not alone, if this is broken, we do have a large customer base who exclusively use only binary telnet for all their service. I have not heard any complains from them from 3.7.24.. krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec
Subject: Re: (usr-tc) DNIS and ANI support in 3.0.2 of the PRI code
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-14 00:08:46
On Sat, 14 Feb 1998, Peter D. Mayer wrote: > I've posted to the list about this before, but never got a response. > > Our boxes USED to get the DNIS and ANI just fine from our PRI's, but ever > since we upgrade the Dual PRI card to 3.0.2, it just disappeared. We still > have one box on the 2.5.3 code, and it grabs them just fine. > Before I go downgrading my PRI cards back to the 2.5.3 code, I'd like to > know if this is a known problem? And if I do have to downgrade, will there > be any compatibility issues with the rest of my equipment which is now on > TCS 3.0.2 code? > Any help on this would be appreciated, as we used to use this information to > track security (with the ANI) and would like to use it in the future to > validate our dedicated customers. There is no know compatibility issues with 3.0.2 code and rest of your equipment. You say here that you upgraded your pri to 3.0.2 - Did you also upgrade your modems? What I suggest is that you open a ticket for this with USR, that will enable us to track this issue and also find out what is wrong. krish > > Oh and one more issue: I would like to upgrade our NMC's to 16MB of ram, > but when I put a 16 meg SIMM in them it only goes up from 4MB to 8MB. Will > they just not accept 16 megs, or is there a jumper you have to set? We put > the same SIMMS in our NETServers to upgrade them, and they worked fine. The > NMC's are hardware revision 3.0. > > Thanks, > Peter D. Mayer > NetWalk Tech Support > dmayer@netwalk.com > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) telnet wierdness
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-14 00:16:26
On Sat, 14 Feb 1998, Brian wrote: > > This is a rather strange post, but because I cannot find an explaination, > I am going to go ahead and post: > > When we telnet to the netservers of any of our hubs, the response is > instantaneous. A login prompt is brought up without any hesitation > whatsoever. > > However, there is one netserver, that when I telnet to it, it get to: > > secrets:~$ telnet usr4ts1 > Trying 208.206.76.41... > Connected to usr4ts1.shreve.net. > Escape character is '^]'. > > then it waits about .75 seconds, and then the rest comes thru: > > U.S. Robotics > Total Control (tm) NETServer Card > > login: > > > I swapped out the ethernet cable. I also changed switch ports, still no > go. > > An inventory reveals that the netserver nac/nic are the same > hardware/software as other hubs I have. > > This hesitation makes me beleave that there are errors or resends or > something to account for the delay, and I am wondering if I should swap > out nics and if that improves it, should I send back this wierd nic. I would try to trace the traffic from the NETServer - see what is happening from the point you try to telnet, as soon as the NETServer receives the TCP request - it will send the sync response - Check to see if you are loosing packets or if the packets are missing in sequence. krish > > Anyone have any ideas for something I should look for in a "show" on the > netservers? > > Brian > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-14 01:06:01
Thus spake Tatai SV Krishnan >On Fri, 13 Feb 1998, Pete Ashdown wrote: >> Jeff Mcadams said once upon a time: >> >> >I just want some decent quality control out of USR on these code >> >releases! How hard is it to check whether a zmodem file transfer >> >works!? >Zmodem does work with Telnet, I am using a linux machine and I use the >command sz -e to start the download. I don't see the problem - What host >are you using? Have you tried with escape code for your zmodem version? If telnet were 8-bit clean (binary mode) as you advertise, I shouldn't *have* to use -e, or at least I've never had to in the past! Regardless, running ppp and uucp over telnet also broke with this upgrade (though as I mentioned in another post, we do have a workaround now) indicating a lack of a binary connection as well. Again, we're running some (old) accounts with ppp on our shell servers, and also have uucp accounts on the shell servers, and were using telnet (previously binary mode) from the netservers to get to them, we've had to come up with another way to do it since 3.7.24 broke the binary mode support in telnet. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) 3.7.24 problem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-14 09:15:54
Thus spake Tatai SV Krishnan >On Sat, 14 Feb 1998, Jeff Mcadams wrote: >> If telnet were 8-bit clean (binary mode) as you advertise, I shouldn't >> *have* to use -e, or at least I've never had to in the past! >> Regardless, running ppp and uucp over telnet also broke with this upgrade >> (though as I mentioned in another post, we do have a workaround now) >> indicating a lack of a binary connection as well. >Telnet is 7 bit, binary telnet is 8-bit. We support both. Not in 3.7.24 you don't. :) >You are not >considering the problems on your host side, if you host side support >binary telnet you need not do -e. Have been able to do this from our host with all the terminal server equipment we've ever run (Xyplex, Livingston, and USR before 3.7.24), when we put 3.7.24 on, sz without the -e quit working...as well as ppp and uucp from hosts over telnet. >What I suggest is not a workarond for our box, it is a workaround for the >host. We tried that since we had source code to telnetd, but couldn't change anything that would fix it...our conclusion was that the telnet on the netserver was broken WRT 8-bit'edness. >> Again, we're running some (old) accounts with ppp on our shell servers, >> and also have uucp accounts on the shell servers, and were using telnet >> (previously binary mode) from the netservers to get to them, we've had >> to come up with another way to do it since 3.7.24 broke the binary mode >> support in telnet. >Atleast from what I have seen so far with 3.7.24 code I have not seen >anything wrong with binary telnet. I will however check it out on >monday and let you know ( just need to test a litte more :-)... ) >If its broken we will fix it. Yup, you need to look farther then, 'cause its definitely broken. :) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) telnet wierdness
From: Brian <signal@shreve.net>
Date: 1998-02-14 10:15:53
This is a rather strange post, but because I cannot find an explaination, I am going to go ahead and post: When we telnet to the netservers of any of our hubs, the response is instantaneous. A login prompt is brought up without any hesitation whatsoever. However, there is one netserver, that when I telnet to it, it get to: secrets:~$ telnet usr4ts1 Trying 208.206.76.41... Connected to usr4ts1.shreve.net. Escape character is '^]'. then it waits about .75 seconds, and then the rest comes thru: U.S. Robotics Total Control (tm) NETServer Card login: I swapped out the ethernet cable. I also changed switch ports, still no go. An inventory reveals that the netserver nac/nic are the same hardware/software as other hubs I have. This hesitation makes me beleave that there are errors or resends or something to account for the delay, and I am wondering if I should swap out nics and if that improves it, should I send back this wierd nic. Anyone have any ideas for something I should look for in a "show" on the netservers? Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) 5ESS & Trunk Groups
From: Brian <signal@shreve.net>
Date: 1998-02-14 10:24:20
We had 12 PRI's that we wanted direct inward dial-in numbers for each two PRI's so we could do testing. Our telco used a 5ESS. They had initially set up 6 trunk groups, of 2 PRI's each, with hunting between them like: 213-4600 <--- lead in number Trunk Group # PRI spans 1 1 & 2 2 3 & 4 3 5 & 6 4 7 & 8 5 9 & 10 6 11 & 12 when you dialed the lead in number, the idea was, TG #1 would roll over to #2, and #2 would roll over to #3, etc. This worked up until span 8 got filled, and it never rolled over to #9 (which is trunk group 5). After an emergency call into the telco, the next day the telco came back and said: "We thought we could do 8 Trunk groups, but the fact is the 5ESS can only do 4 trunk groups for a lead in number, and so we cant do this" Now, I know some of you have DID numbers on your spans for testing, I know Jesse Sipperel(sp?) does and he is on a 5ESS. What I would like to know is: 1. Is my telco full of it, and this can be done with trunk groups? If so whats the limit? 2. Maybe this cant be done with Trunk Groups and hunting, but how is it done? 3. Are any of you on 5ESS switches, with more than 4 "test" DID numbers for your spans, broken into more than 4 groups? Thanks, Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) DNIS and ANI support in 3.0.2 of the PRI code
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-14 10:58:30
I've posted to the list about this before, but never got a response. Our boxes USED to get the DNIS and ANI just fine from our PRI's, but ever since we upgrade the Dual PRI card to 3.0.2, it just disappeared. We still have one box on the 2.5.3 code, and it grabs them just fine. Before I go downgrading my PRI cards back to the 2.5.3 code, I'd like to know if this is a known problem? And if I do have to downgrade, will there be any compatibility issues with the rest of my equipment which is now on TCS 3.0.2 code? Any help on this would be appreciated, as we used to use this information to track security (with the ANI) and would like to use it in the future to validate our dedicated customers. Oh and one more issue: I would like to upgrade our NMC's to 16MB of ram, but when I put a 16 meg SIMM in them it only goes up from 4MB to 8MB. Will they just not accept 16 megs, or is there a jumper you have to set? We put the same SIMMS in our NETServers to upgrade them, and they worked fine. The NMC's are hardware revision 3.0. Thanks, Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com
Subject: Re: (usr-tc) Radius crashes
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-14 11:12:51
On Fri, 13 Feb 1998, MegaZone wrote: > Once upon a time Mike shaped the electrons to say... > >I think this is a bug in Livingston radius.. When a NS reboots it sends a > > It is odd - since an Accounting-On packet doesn't make us reboot. It is > just logged as: > Acct-Status-Type = 7 > > It should not cause a reboot, and I've seen this be logged and there are > many users running TCs against Lucent RADIUS without reboots. Version 2.0 of Livingston Radius would stop responding whenever my TC was rebooted. There was a bug in the code that caused Radius to die whenever a unknown packet type was sent. I patched a few lines of code, and my Radius server has never died again. Brian
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Charles Hill <chill@ionet.net>
Date: 1998-02-14 11:29:00
We have run into that same limitation and the solution was to sacrifice a channel on each PRI we wanted to test dial. . . and they don't go to waste because employees use them. -CH On Sat, 14 Feb 1998, Brian wrote: > We had 12 PRI's that we wanted direct inward dial-in numbers for each two > PRI's so we could do testing. > > Our telco used a 5ESS. > > They had initially set up 6 trunk groups, of 2 PRI's each, with hunting > between them like: > > 213-4600 <--- lead in number > > Trunk Group # PRI spans > --------------------------------- > 1 1 & 2 > 2 3 & 4 > 3 5 & 6 > 4 7 & 8 > 5 9 & 10 > 6 11 & 12 > > > when you dialed the lead in number, the idea was, TG #1 would roll over to > #2, and #2 would roll over to #3, etc. > > This worked up until span 8 got filled, and it never rolled over to #9 > (which is trunk group 5). After an emergency call into the telco, the > next day the telco came back and said: > > "We thought we could do 8 Trunk groups, but the fact is the 5ESS can only > do 4 trunk groups for a lead in number, and so we cant do this" > > Now, I know some of you have DID numbers on your spans for testing, I know > Jesse Sipperel(sp?) does and he is on a 5ESS. What I would like to know > is: > > 1. Is my telco full of it, and this can be done with trunk groups? If so > whats the limit? > > 2. Maybe this cant be done with Trunk Groups and hunting, but how is it > done? > > 3. Are any of you on 5ESS switches, with more than 4 "test" DID numbers > for your spans, broken into more than 4 groups? > > Thanks, > > Brian > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-14 11:35:28
On Sat, 14 Feb 1998, Brian wrote: > when you dialed the lead in number, the idea was, TG #1 would roll over to > #2, and #2 would roll over to #3, etc. > > This worked up until span 8 got filled, and it never rolled over to #9 > (which is trunk group 5). After an emergency call into the telco, the > next day the telco came back and said: > > "We thought we could do 8 Trunk groups, but the fact is the 5ESS can only > do 4 trunk groups for a lead in number, and so we cant do this" We wanted to do a similiar thing on a 5ESS, but our result was far worse. Our main trunk group has 6 CT1 spans, and we wanted to pull one span out of the main trunk group, and put it in a seperate trunk group, and the main trunk would overflow to the new trunk group when busy. I was all set to have this change done one day this week, so I called the tech and told him to go ahead. The tech then told me that he couldn't do it as they had tried it with another ISP, and only the first 4 DS0s of the second trunk group would fill before the switch started returning busies. The other ISP had set up a seperate trunk group for each chassis they have, but they had to have one large trunk group set up so that they could fill all of their DS0s. Brian
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Brian <signal@shreve.net>
Date: 1998-02-14 12:13:37
On Sat, 14 Feb 1998, Charles Hill wrote: > > We have run into that same limitation and the solution was to sacrifice a > channel on each PRI we wanted to test dial. . . and they don't go to waste > because employees use them. -CH > > On Sat, 14 Feb 1998, Brian wrote: Ok, so let me get this straight, you can make "test" numbers into your PRI's, but that test number is not part of the actual hunt group? /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) telnet wierdness
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-14 14:07:33
This was brought up on the list a while back, and should be normal behavior. It takes that amount of time for the netserver to set things up for the first connect; subsequent telnet connections should be instantanous (if you hold at least one open). - lv On Sat, 14 Feb 1998, Brian wrote: > > This is a rather strange post, but because I cannot find an explaination, > I am going to go ahead and post: > > When we telnet to the netservers of any of our hubs, the response is > instantaneous. A login prompt is brought up without any hesitation > whatsoever. > > However, there is one netserver, that when I telnet to it, it get to: > > secrets:~$ telnet usr4ts1 > Trying 208.206.76.41... > Connected to usr4ts1.shreve.net. > Escape character is '^]'. > > then it waits about .75 seconds, and then the rest comes thru: > > U.S. Robotics > Total Control (tm) NETServer Card > > login: > > > I swapped out the ethernet cable. I also changed switch ports, still no > go. > > An inventory reveals that the netserver nac/nic are the same > hardware/software as other hubs I have. > > This hesitation makes me beleave that there are errors or resends or > something to account for the delay, and I am wondering if I should swap > out nics and if that improves it, should I send back this wierd nic. > > Anyone have any ideas for something I should look for in a "show" on the > netservers? > > Brian > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Doug McClure <dmcclure@infi.net>
Date: 1998-02-14 14:31:38
Yes, ask for a ZTN on each PRI. It's just an extra number for the PRI. Doug At 12:13 PM 2/14/98 -0600, you wrote: >On Sat, 14 Feb 1998, Charles Hill wrote: > >> >> We have run into that same limitation and the solution was to sacrifice a >> channel on each PRI we wanted to test dial. . . and they don't go to waste >> because employees use them. -CH >> >> On Sat, 14 Feb 1998, Brian wrote: > >Ok, so let me get this straight, you can make "test" numbers into your >PRI's, but that test number is not part of the actual hunt group? > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Network Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > >
Subject: Re: (usr-tc) telnet wierdness
From: Brian <signal@shreve.net>
Date: 1998-02-14 17:56:40
On Sat, 14 Feb 1998, Laszlo Vecsey wrote: > This was brought up on the list a while back, and should be normal > behavior. It takes that amount of time for the netserver to set things up > for the first connect; subsequent telnet connections should be > instantanous (if you hold at least one open). > > - lv > laszlo, What your describing is just things like dns name getting cached, arp cache being updated, etc, etc, so that after the first connect things are faster, this is not the case here. This is something that is wrong, and I can connect 100 times, and have 100 hesitations. I think I will tcpdump it. Everything is etherswitched so tcpdumping is kinda difficult. I think I can configure a port of our Catalyst 3200 as a "probe" port, and actually see *all* traffic. I will just do that and go in there and sniff anything with a source or destination of the netserver. Brian > On Sat, 14 Feb 1998, Brian wrote: > > > > > This is a rather strange post, but because I cannot find an explaination, > > I am going to go ahead and post: > > > > When we telnet to the netservers of any of our hubs, the response is > > instantaneous. A login prompt is brought up without any hesitation > > whatsoever. > > > > However, there is one netserver, that when I telnet to it, it get to: > > > > secrets:~$ telnet usr4ts1 > > Trying 208.206.76.41... > > Connected to usr4ts1.shreve.net. > > Escape character is '^]'. > > > > then it waits about .75 seconds, and then the rest comes thru: > > > > U.S. Robotics > > Total Control (tm) NETServer Card > > > > login: > > > > > > I swapped out the ethernet cable. I also changed switch ports, still no > > go. > > > > An inventory reveals that the netserver nac/nic are the same > > hardware/software as other hubs I have. > > > > This hesitation makes me beleave that there are errors or resends or > > something to account for the delay, and I am wondering if I should swap > > out nics and if that improves it, should I send back this wierd nic. > > > > Anyone have any ideas for something I should look for in a "show" on the > > netservers? > > > > Brian > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) telnet wierdness
From: Brian <signal@shreve.net>
Date: 1998-02-14 17:56:40
On Sat, 14 Feb 1998, Laszlo Vecsey wrote: > This was brought up on the list a while back, and should be normal > behavior. It takes that amount of time for the netserver to set things up > for the first connect; subsequent telnet connections should be > instantanous (if you hold at least one open). > > - lv > laszlo, What your describing is just things like dns name getting cached, arp cache being updated, etc, etc, so that after the first connect things are faster, this is not the case here. This is something that is wrong, and I can connect 100 times, and have 100 hesitations. I think I will tcpdump it. Everything is etherswitched so tcpdumping is kinda difficult. I think I can configure a port of our Catalyst 3200 as a "probe" port, and actually see *all* traffic. I will just do that and go in there and sniff anything with a source or destination of the netserver. Brian > On Sat, 14 Feb 1998, Brian wrote: > > > > > This is a rather strange post, but because I cannot find an explaination, > > I am going to go ahead and post: > > > > When we telnet to the netservers of any of our hubs, the response is > > instantaneous. A login prompt is brought up without any hesitation > > whatsoever. > > > > However, there is one netserver, that when I telnet to it, it get to: > > > > secrets:~$ telnet usr4ts1 > > Trying 208.206.76.41... > > Connected to usr4ts1.shreve.net. > > Escape character is '^]'. > > > > then it waits about .75 seconds, and then the rest comes thru: > > > > U.S. Robotics > > Total Control (tm) NETServer Card > > > > login: > > > > > > I swapped out the ethernet cable. I also changed switch ports, still no > > go. > > > > An inventory reveals that the netserver nac/nic are the same > > hardware/software as other hubs I have. > > > > This hesitation makes me beleave that there are errors or resends or > > something to account for the delay, and I am wondering if I should swap > > out nics and if that improves it, should I send back this wierd nic. > > > > Anyone have any ideas for something I should look for in a "show" on the > > netservers? > > > > Brian > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Brian <signal@shreve.net>
Date: 1998-02-14 17:57:45
On Sat, 14 Feb 1998, Doug McClure wrote: > > Yes, ask for a ZTN on each PRI. It's just an extra number for the PRI. > > Doug And you lose the ZTN in the hunt right? so for each PRI you have a ZTN for, you have 22 channels that can fall under your normal lead-in hunt number correct? Brian > > At 12:13 PM 2/14/98 -0600, you wrote: > >On Sat, 14 Feb 1998, Charles Hill wrote: > > > >> > >> We have run into that same limitation and the solution was to sacrifice a > >> channel on each PRI we wanted to test dial. . . and they don't go to waste > >> because employees use them. -CH > >> > >> On Sat, 14 Feb 1998, Brian wrote: > > > >Ok, so let me get this straight, you can make "test" numbers into your > >PRI's, but that test number is not part of the actual hunt group? > > > >/-------------------------- signal@shreve.net -----------------------------\ > >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > >| Network Administrator | Perl, Linux | Web hosting, online stores, | > >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Radius crashes
From: MegaZone <megazone@megazone.org>
Date: 1998-02-14 18:11:59
Once upon a time Brian Elfert shaped the electrons to say... >Version 2.0 of Livingston Radius would stop responding whenever my TC >was rebooted. Do you now if that is still true in 2.0.1? 2.0 had a few problems, but those that were reported before 2.0.1 should have been fixed. If it is still in 2.0.1 please send your patch to cdr@livingston.com - RADIUS 2.1 is well along now and they've made an effort to fix anything reported from 2.0.1. -MZ -- <URL:mailto:megazone@megazone.org> Gweep, author, webmaster, human being, me "A little nonsense now and then, is relished by the wisest men" 510-527-0944 <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia!
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Charles Hill <chill@ionet.net>
Date: 1998-02-14 18:42:31
On Sat, 14 Feb 1998, Brian wrote: > On Sat, 14 Feb 1998, Doug McClure wrote: > > > Yes, ask for a ZTN on each PRI. It's just an extra number for the PRI. > > > > Doug > > And you lose the ZTN in the hunt right? so for each PRI you have a ZTN > for, you have 22 channels that can fall under your normal lead-in hunt > number correct? I have never heard it referred to as a ZTN, but you are correct; 22 channels are in your main hunt group, one is set aside with a separate number. It's a Lucent 5E thing. Nortel DMS switches don't have the six trunk group hunt limitation and test numbers are no problem. One good side-effect was that I could lose a quad card and still have enough modems for the 44 channels. -CH
Subject: Re: (usr-tc) Radius
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-14 21:19:30
On Sun, 15 Feb 1998, uri simhon wrote: > Hi !!! > I have netserver 8 ver 3.2.5.3 and I have Security/Accounting Server > software ver 4.3 . > I install this software on Nt server 4.00 > I want to do the authentication from the Nt server and it is not working > . > Can you help me please . Uri, The NETServer 8/16 is just like the NETServer. Make sure the following is set set s1-s8 security on set authen <ip address of the Radius Server> set secret <secret key> reset all This will work. If you have problems with this you will see the log on the Radius server where it will say what is happening. good luck ...:-)... krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec > > Thanks > Uri Simhon > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) telnet wierdness
From: Brian <signal@shreve.net>
Date: 1998-02-14 22:46:27
On Sun, 15 Feb 1998, Bob Purdon wrote: > > > When we telnet to the netservers of any of our hubs, the response is > > instantaneous. A login prompt is brought up without any hesitation > > whatsoever. > > I observed something interesting, and possibly related, the other day > while tinkering with RIPv2. Without RIPv2 I was getting the same 1 > second delay that you're seeing. With RIPv2 enabled it was instantaneous. > Turn RIPv2 off again and back came the delay. > > I know you run RIPv2. Is this one chassis running plain old RIPv1 > perhaps? Its runing RIPv2, but the TC's still respond to version 1 packets in v2 mode, which I have no idea why, I have a ticket open with 3com on this. I thought, that with "set ripv2 on", that the netserver would do ripv2 and *not* v1, but apparently, this setting enables both. I am curious about what you speak of though, doesn't sound right..........I can't wait until ospf is out. > > Regards, > > Bob Purdon, > Technical Manager, > Southern Internet Services. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Doug McClure <dmcclure@infi.net>
Date: 1998-02-14 22:58:08
At 05:57 PM 2/14/98 -0600, you wrote: >On Sat, 14 Feb 1998, Doug McClure wrote: > >> >> Yes, ask for a ZTN on each PRI. It's just an extra number for the PRI. >> >> Doug > >And you lose the ZTN in the hunt right? so for each PRI you have a ZTN >for, you have 22 channels that can fall under your normal lead-in hunt >number correct? No, it's just another number at the switch. Everything happens on the D channel, you'll never loose a B channel for the use of a number. The ZTN is Bell Atlantic's code for 'an extra number' is all. I've got a ton of them on different PRI's. We usually get one when we add service to an existing hunt so we can test the circuit without having to wait for calls to roll into it from the existing hunt. It's all a matter of provisioning. Doug
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Doug McClure <dmcclure@infi.net>
Date: 1998-02-14 23:00:42
At 06:42 PM 2/14/98 -0600, you wrote: >On Sat, 14 Feb 1998, Brian wrote: > >> On Sat, 14 Feb 1998, Doug McClure wrote: >> >> > Yes, ask for a ZTN on each PRI. It's just an extra number for the PRI. >> > >> > Doug >> >> And you lose the ZTN in the hunt right? so for each PRI you have a ZTN >> for, you have 22 channels that can fall under your normal lead-in hunt >> number correct? > >I have never heard it referred to as a ZTN, but you are correct; 22 >channels are in your main hunt group, one is set aside with a separate >number. It's a Lucent 5E thing. Nortel DMS switches don't have the six >trunk group hunt limitation and test numbers are no problem. > >One good side-effect was that I could lose a quad card and still have >enough modems for the 44 channels. > >-CH That's wrong. All the signalling and control on a PRI ALWAYS occurs on the D channel. I could have 23 different numbers point to the same PRI it wouldn't matter (I'm guessing, there 'could' be a limit). I've got some PRIs with 5 different numbers pointed to it. I still get 23 available B channels on that PRI. Doug
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Charles Hill <chill@ionet.net>
Date: 1998-02-14 23:15:22
> >> And you lose the ZTN in the hunt right? so for each PRI you have a ZTN > >> for, you have 22 channels that can fall under your normal lead-in hunt > >> number correct? > > > >I have never heard it referred to as a ZTN, but you are correct; 22 > >channels are in your main hunt group, one is set aside with a separate > >number. It's a Lucent 5E thing. Nortel DMS switches don't have the six > >trunk group hunt limitation and test numbers are no problem. > > > >One good side-effect was that I could lose a quad card and still have > >enough modems for the 44 channels. > > That's wrong. All the signalling and control on a PRI ALWAYS occurs on the > D channel. I could have 23 different numbers point to the same PRI it > wouldn't matter (I'm guessing, there 'could' be a limit). I've got some > PRIs with 5 different numbers pointed to it. I still get 23 available B > channels on that PRI. I have 23 available channels, just 22 remain in the large trunk group. If it's possible to point a test number to a span that is in the middle of a 40-span hunt group without breaking off a channel and creating a separate trunk group, then I'm working with 5E switch technicians who don't know how. Please introduce me to a switch tech who can school them. -CH
Subject: (usr-tc) T1s, BellSouth, and Half-Busies
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-14 23:18:53
Today was fun; I soft busied some span lines last night, and went in this morning to rework their cabling and upgrade the firmware on the connected units. Suddenly, our rollover was rolling-over idle channels and one whole T1 -- the fourth of ten in our hunt group -- was awol. BellSouth's trouble-reporting number was broken (!) so my boss pulled out his secret list of beepers and home phone numbers. It was about three hours after we started trying to get somebody to look at our systems that they dispatched a senior engineer for our town; the poor guy lives an hour away, so I waited for a while. It turned out that something triggered by me taking down the chanellized T1s temporarily had made the channel modules (?) at the Central Office cut and run, making the switch believe that they were busy. When Larry Cook, who had driven in to work on it, manually released them at the switch, a call could come through, but then it could get stuck again. He studied the engineering notes, and found that several different brands of channel modules had been used, and that in one of them some odd configuration option had been chosen. He fixed that, and that seems to have fixed the half-busy problem. I'm not sure what he did to fix the whole span that was out of whack, but Larry's been known to diagnose wiring problems inside the CO while on the phone with me. Anyway, we're back up and running, and I have renewed faith that there are very good people out there on the other end of the span line. :) --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: Re: (usr-tc) T1s, BellSouth, and Half-Busies
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-15 00:11:10
Thus spake Mark R. Lindsey >Anyway, we're back up and running, and I have renewed faith that there >are very good people out there on the other end of the span line. :) Just to put my $.02 in on BellSouth, being one of the major complainers about them. My experience was almost always that once you got someone significant no the other end of the line, you could almost always find your problem very quickly. My main problem is that it can sometimes be nearly impossible to get someone significant on the other end of the line. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) routing problem
From: Josh Richards <jrichard@livingston.com>
Date: 1998-02-15 00:13:12
On 12 Feb 1998, Brian wrote: > On Thu, 12 Feb 1998, Charles Sprickman wrote: > > > Where does the Cisco seem to be getting this route? Is there a > > ripquery-like command available there? Do you have "ip subnet zero" > > enabled on the cisco? > > > > Charles > > Yes ip subnet-zero is enabled on the router. The router is learning this > route from the Total Control, as they are the only things listed as > sending updates. [..] > > So I did ripquery, and the only thing that comes back with 208.214.45.0 > and that /24 netmask, is the cisco. The cisco knows nothing about > 208.214.45.0 however, it must have learned it from the TC hub. On the Cisco do a "show ip route 208.214.45.0"....it should tell you exactly the route it has and how, when, what protocol, and where it is actually receiving the route from. [..] > I wish I could better track this down. ---- Josh Richards - <jrichard@livingston.com> - [Beta Engineer] LUCENT Technologies - Remote Access Business Unit (formerly Livingston Enterprises, Inc.) http://www.livingston.com/
Subject: Re: (usr-tc) Multi PPP
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-15 01:23:24
Uri, change your modem code on the NETServer 8/16. krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec On Sun, 15 Feb 1998, uri simhon wrote: > Hi !!! > I have netserver 8/I ver 3.2.5.3 modem's ver 1.5.2 . > I try to work multi ppp (128k) and it is not working . > I did upgrade to the modems to code 2.5.8 and still I don't get multi > ppp , the modem I called from it is courier I it open 2 channels but > after that it drop the other channel . > Just to let you know I call from tihs modem to Total control (pri) and I > am able to do multi ppp and it is working fine . > may someone help me ? > > Thanks > Uri Simhon > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) T1s, BellSouth, and Half-Busies
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-15 01:36:51
At 12:11 AM 2/15/98 -0500, Jeff Mcadams wrote: >Thus spake Mark R. Lindsey >>Anyway, we're back up and running, and I have renewed faith that there >>are very good people out there on the other end of the span line. :) > >Just to put my $.02 in on BellSouth, being one of the major complainers >about them. My experience was almost always that once you got someone >significant no the other end of the line, you could almost always find >your problem very quickly. My main problem is that it can sometimes be >nearly impossible to get someone significant on the other end of the >line. This list is not the proper forum for a discussion of BellSouth service (may their slimy souls rod in hell). If there is one, please let me know about it, because I do love to bash them after our most recent HellSouth nightmare. In this nightmare, the owner of the isp stays up late calling random numbers throughout his exchange and receiving busy signals on each and every one. He tries calling from different locations in town.. All 675-XXXX numbers are busy! So he tries calling his reseller, a clec, to report the problem. Busy.. A hundred and fify times, busy... Sound unbelieable? A bad dream perhaps.. This nightmare was real. With around 120 channels of 161 available, 9 of 10 calls during peak resulted in a busy signal for nearly 3 weeks until they added more trunks between various co's and switches. And HellSouth told me in november that they could sell me no more pri's until february.. It was bad, even the fire and police stations were affected. HellSouth got on the news and blamed it on the internet of course. (cowards) Lucky we are still in business. In the meantime, they see fit to enclose in their phone bills an advertise for bellsouth.net with a guarantee of no busy signals!! And they refuse an invitation to even talk to us because we use a clec as an "agent" to resell us Bellsouth service at 10% off.. Yeah with hosting lists being the fashion, I should start bellsouth-sucks-bad@shreve.net.. But I don't want to look bad just in case I can sue them.. allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Changing radius ports?
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-15 07:25:41
On Sun, 15 Feb 1998, Support wrote: > The Radius Ver in SET AUTHENTIC <IpAddr> <port> <secret> <Radius Ver> > versions are in the in the parametric reference but no where can I see the > syntax of how to set it to group. Do you know where I can find it's values. > Especial to enable the use of Radius 2.01 accounting. set authen 207.24.169.174/1645 secret v2 But remember Radius 2.01 accounting has nothing to do with radius version 2. These are totally different krish > > Thanks > > ---------- > > From: Jason_Kelton@3com.com > > To: usr-tc@lists.xmission.com > > Subject: Re: (usr-tc) Changing radius ports? > > Date: Thursday, February 12, 1998 10:05 PM > > > > Yep... use the following command line > > > > SET AUTHENTIC <IpAddr> <port> <secret> <Radius Ver> > > I think a forward slash is required between the IP and the port > > 123.456.789.xxx/1600 > > > > I believe its documented in the NETServer manual > > > > Regards, > > > > Jason. > > > > > > > > > > phil@freed.com on 14/02/98 01:50:18 > > > > Please respond to usr-tc@lists.xmission.com > > > > To: usr-tc@xmission.com > > cc: (bcc: Jason Kelton/AU/3Com) > > Subject: (usr-tc) Changing radius ports? > > > > > > > > > > > > > > > > Is there a way to change the standard Radius authentication and > > accounting ports in a TC Netserver card? > > Thanks.... > > --phil > > "All my life, I always wanted to be somebody. > > Now I see that I should have been more specific." > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) telnet wierdness
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-15 08:06:14
On Sat, 14 Feb 1998, Brian wrote: > On Sat, 14 Feb 1998, Laszlo Vecsey wrote: > > > This was brought up on the list a while back, and should be normal > > behavior. It takes that amount of time for the netserver to set things up > > for the first connect; subsequent telnet connections should be > > instantanous (if you hold at least one open). > > > > - lv > > > laszlo, > > What your describing is just things like dns name getting cached, arp > cache being updated, etc, etc, so that after the first connect things are > faster, this is not the case here. This is something that is wrong, and I > can connect 100 times, and have 100 hesitations. I think I will tcpdump > it. > Did you try this 100 times in sequence (i.e, connecting with some utility, disconnecting, and then connecting again?) If you try a couple in parallel you may notice they start to connect instantly. This is the behavior I've noticed, and the same behavior a USR technician has confirmed on this email list before.. it takes time for the netserver to set things up for the initial telnet connection, and thats where the delay before the login: prompt comes from. However, if you stay logged in with at least one telnet session to the netserver, subsequent telnets to the netserver should bring up the login prompt instantly. Disconnect with all of them, and the delay will be back again with your first connect. - lv > > On Sat, 14 Feb 1998, Brian wrote: > > > > > > > > This is a rather strange post, but because I cannot find an explaination, > > > I am going to go ahead and post: > > > > > > When we telnet to the netservers of any of our hubs, the response is > > > instantaneous. A login prompt is brought up without any hesitation > > > whatsoever. > > > > > > However, there is one netserver, that when I telnet to it, it get to: > > > > > > secrets:~$ telnet usr4ts1 > > > Trying 208.206.76.41... > > > Connected to usr4ts1.shreve.net. > > > Escape character is '^]'. > > > > > > then it waits about .75 seconds, and then the rest comes thru: > > > > > > U.S. Robotics > > > Total Control (tm) NETServer Card > > > > > > login: > > > > > > > > > I swapped out the ethernet cable. I also changed switch ports, still no > > > go. > > > > > > An inventory reveals that the netserver nac/nic are the same > > > hardware/software as other hubs I have. > > > > > > This hesitation makes me beleave that there are errors or resends or > > > something to account for the delay, and I am wondering if I should swap > > > out nics and if that improves it, should I send back this wierd nic. > > > > > > Anyone have any ideas for something I should look for in a "show" on the > > > netservers? > > > > > > Brian > > > > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > > > > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the message. > > > For information on digests or retrieving files and old messages send > > > "help" to the same address. Do not use quotes in your message. > > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) TEST - ignore this
From: Ken Hunter, Aspiring Technologies <ken@aspire.net>
Date: 1998-02-15 08:53:44
test post. not sure if I was getting out. - ************************************************************************ Web Hosting, E-mail addresses, DNS services, Dedicated connectivity. 33.6, 56k, ISDN Dialup coming soon to Prince William County VA. Ken Hunter Aspiring Technologies mailto:ken@aspire.net 9304 Troy Drive http://www.aspire.net Nokesville, Va 20181 ************************************************************************
Subject: (usr-tc) Radius
From: uri simhon <uri@team.co.il>
Date: 1998-02-15 09:54:17
Hi !!! I have netserver 8 ver 3.2.5.3 and I have Security/Accounting Server software ver 4.3 . I install this software on Nt server 4.00 I want to do the authentication from the Nt server and it is not working . Can you help me please . Thanks Uri Simhon
Subject: (usr-tc) new o/s on netserver 8i-plus
From: Ken Hunter, Aspiring Technologies <ken@aspire.net>
Date: 1998-02-15 10:58:26
I just setup a new usr netserver 8/i os version: 4.0.13 imodem flash: 2.2.2 radius: 2.0b11 (lucent version) 1. I don't think I like the new o/s - it appears to be unstable. 2. I turn RIP off, save all, reboot, it returns as being enabled. I don't use rip currently. Therefore, I don't want it turned on. I may need it later - but not now. I have ip forwarding turned on - and that's it. Am I missing something? 3. The port limit or concurrent login limit doesn't seem to work either. I can set the Port-Limit param in the lucent radius file for a user to say 1, but can login more than once for that user. 4. I'm told the earlier os versions were more like the lucent ComOs - which I am more used to. Yes - I know not be closed minded to a new way of doing things - but the new way needs to not be flakey. Are there any O/S versions more stable? If so, can I get this code and roll my netserver back to this o/s? I could care less about the newest o/s - I am more interested in a setup that won't fly apart under fire. 5. As it sits currently, I wouldn't open the doors to dial in service. These folks around here would skin me alive with the way this thing behaves. Thanks in advance, Ken :) - ************************************************************************ Web Hosting, E-mail addresses, DNS services, Dedicated connectivity. 33.6, 56k, ISDN Dialup coming soon to Prince William County VA. Ken Hunter Aspiring Technologies mailto:ken@aspire.net 9304 Troy Drive http://www.aspire.net Nokesville, Va 20181 ************************************************************************
Subject: Re: (usr-tc) routing problem
From: Brian <signal@shreve.net>
Date: 1998-02-15 11:33:21
On Sun, 15 Feb 1998, Josh Richards wrote: > On 12 Feb 1998, Brian wrote: > > > On Thu, 12 Feb 1998, Charles Sprickman wrote: > > > > > Where does the Cisco seem to be getting this route? Is there a > > > ripquery-like command available there? Do you have "ip subnet zero" > > > enabled on the cisco? > > > > > > Charles > > > > Yes ip subnet-zero is enabled on the router. The router is learning this > > route from the Total Control, as they are the only things listed as > > sending updates. > [..] > > > > So I did ripquery, and the only thing that comes back with 208.214.45.0 > > and that /24 netmask, is the cisco. The cisco knows nothing about > > 208.214.45.0 however, it must have learned it from the TC hub. > > On the Cisco do a "show ip route 208.214.45.0"....it should tell you > exactly the route it has and how, when, what protocol, and where it is > actually receiving the route from. O, thats the best part, check this out: R 208.214.45.0/28 [120/1] via 208.206.76.35, 00:00:28, Ethernet0 BUT, low and behold! mercury:~$ /etc/ripquery -r stargate 84 bytes from stargate.shreve.net(208.206.76.1) version 2: 0.0.0.0/0.0.0.0 router 0.0.0.0 metric 1 208.232.62.0/255.255.255.0 router 0.0.0.0 metric 2 208.214.44.0/255.255.255.0 router 208.206.76.35 metric 2 208.214.45.0/255.255.255.0 router 0.0.0.0 metric 2 tell me that aint screwed up! > > [..] > > I wish I could better track this down. > > ---- > Josh Richards - <jrichard@livingston.com> - [Beta Engineer] > LUCENT Technologies - Remote Access Business Unit > (formerly Livingston Enterprises, Inc.) > http://www.livingston.com/ > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) telnet wierdness
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-15 12:17:46
> When we telnet to the netservers of any of our hubs, the response is > instantaneous. A login prompt is brought up without any hesitation > whatsoever. I observed something interesting, and possibly related, the other day while tinkering with RIPv2. Without RIPv2 I was getting the same 1 second delay that you're seeing. With RIPv2 enabled it was instantaneous. Turn RIPv2 off again and back came the delay. I know you run RIPv2. Is this one chassis running plain old RIPv1 perhaps? Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: Re: (usr-tc) Changing radius ports?
From: Support <mhamrich@drfast.net>
Date: 1998-02-15 14:45:17
The Radius Ver in SET AUTHENTIC <IpAddr> <port> <secret> <Radius Ver> versions are in the in the parametric reference but no where can I see the syntax of how to set it to group. Do you know where I can find it's values. Especial to enable the use of Radius 2.01 accounting. Thanks
Subject: Re: (usr-tc) T1s, BellSouth, and Half-Busies
From: William Behrens <wbehrens@feist.com>
Date: 1998-02-15 16:23:36
From: "William Behrens" <wbehrens@feist.com> Subject: Re: (usr-tc) T1s, BellSouth, and Half-Busies Date: 15 Feb 1998 16:23:36 -0600
Subject: Re: (usr-tc) 5ESS & Trunk Groups
From: Doug McClure <dmcclure@infi.net>
Date: 1998-02-15 17:25:18
I'll see if I can find out, but I recommend talking with provisioning ENGINEERS and not switch TECHS. There's a very big difference! Doug At 11:15 PM 2/14/98 -0600, you wrote: >> >> And you lose the ZTN in the hunt right? so for each PRI you have a ZTN >> >> for, you have 22 channels that can fall under your normal lead-in hunt >> >> number correct? >> > >> >I have never heard it referred to as a ZTN, but you are correct; 22 >> >channels are in your main hunt group, one is set aside with a separate >> >number. It's a Lucent 5E thing. Nortel DMS switches don't have the six >> >trunk group hunt limitation and test numbers are no problem. >> > >> >One good side-effect was that I could lose a quad card and still have >> >enough modems for the 44 channels. >> >> That's wrong. All the signalling and control on a PRI ALWAYS occurs on the >> D channel. I could have 23 different numbers point to the same PRI it >> wouldn't matter (I'm guessing, there 'could' be a limit). I've got some >> PRIs with 5 different numbers pointed to it. I still get 23 available B >> channels on that PRI. > >I have 23 available channels, just 22 remain in the large trunk group. If >it's possible to point a test number to a span that is in the middle of a >40-span hunt group without breaking off a channel and creating a separate >trunk group, then I'm working with 5E switch technicians who don't know >how. Please introduce me to a switch tech who can school them. > >-CH > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > >
Subject: Re: (usr-tc) T1s, BellSouth, and Half-Busies
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-15 18:50:48
At 04:23 PM 2/15/98 -0600, William Behrens wrote: >(Andy Rooney mode on) > >Isn't it funny how its called the Public Telephone Switch Network, yet when >you try to find out why its managed so poorly your told by your RBOC that >its none of your business. If I were to compare the monopoly that Microsoft >has on the software industry to the telephony monopoly that the RBOC's have >on our industry (at least from the inbound side) I think it would be safe >to say internet users would benefit from the DOJ going after the RBOC's >more than Microsoft. > >(Andy Rooney mode off) I think we just need to "pay the man" like bell probably does.. Maybe the isp-c could cut a check to the right congressman or senator.. Or take him hunting or golfing, or maybe to the superbowl.. >Have any of you heard about the congressman who wants ISP's to pay into the >Universal Service fund. He thinks we get a free ride (of course the RBOC's >are applauding this action). How much do we pay for PRI's a month that work >about 1/2 the time. Have you ever asked for credits from your RBOC, because >of a PRI being down for a week ? Ok I know this is not the place for this >so I'll shut up......but I do feel so much better for venting my spleen on >this issue. > This thread should be killed here but I wonder where is the proper forum for much needed venting on this issue??.. inet-access? I subscribe to isp-clec and it's not so great. alot of fluff and no meat.. are there any rboc or telco related lists out there?.. isp's are the last folks on the planet that should be abused with all their skills and resources where global communication is concerned.. am
Subject: (usr-tc) Multi PPP
From: uri simhon <uri@team.co.il>
Date: 1998-02-15 19:31:12
Hi !!! I have netserver 8/I ver 3.2.5.3 modem's ver 1.5.2 . I try to work multi ppp (128k) and it is not working . I did upgrade to the modems to code 2.5.8 and still I don't get multi ppp , the modem I called from it is courier I it open 2 channels but after that it drop the other channel . Just to let you know I call from tihs modem to Total control (pri) and I am able to do multi ppp and it is working fine . may someone help me ? Thanks Uri Simhon
Subject: Re: (usr-tc) telnet wierdness
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-16 10:00:49
> Did you try this 100 times in sequence (i.e, connecting with some > utility, disconnecting, and then connecting again?) OK, I've just put 3.7.24 on one of our chassis and the behaviour has changed. Previously at 3.6.28 and 3.5.33, when using RIPv1 routing, I had a 1 second delay on connecting. Changing to RIPv2 made connections instantaneous. At 3.7.24 it's instantaneous regardless :-) Cheers, Bob.
Subject: (usr-tc) V.90
From: Terry Kennedy <terry@olypen.com>
Date: 1998-02-16 11:28:40
I posted this question a week or so ago and got a response but am still unclear on a certain point. Will the new standard (v.90) work with unmodified versions of X2 modems? KFlex modems? In other works, when the code becomes available and we load it on the netservers, will the customers be forced to upgrade their modems to v.90 before they get a 56k connection again? Although it's conceivable that most customers would upgrade eventually, it would be a nightmare to load the code one night and anwer the phones the next day and start explaining to them that they need to upgrade even if it is just a software solution. Not to mention the ones that need a hardware upgrade. Any one care to shed some light on this. There is nothing at totalservice.com that helps to explain this. Terry Kennedy OlyPen, Inc.
Subject: Re: (usr-tc) PPP/Slip Script problems
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-16 12:07:18
: For customers using trumpet winsock, there are looking for the word "to". : Works for netserver only. One of the problem is telling the customer to : change his/her script. I wonder why 3Com changed it... Remember that recent versions of Trumpet Winsock can use PAP, and thus don't need much of a login script.
Subject: Re: (usr-tc) Changing radius ports?
From: jason_kelton@3com.com
Date: 1998-02-16 12:08:27
hmmm, I believe the Netserver Manual is incorrectly documented. You need to type v1 or v2 for Radius V1 or 2. The Manual says 1 or 2 - this does not work. Regards, Jason Kelton mhamrich@drfast.net on 16/02/98 05:45:17 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) The Radius Ver in SET AUTHENTIC <IpAddr> <port> <secret> <Radius Ver> versions are in the in the parametric reference but no where can I see the syntax of how to set it to group. Do you know where I can find it's values. Especial to enable the use of Radius 2.01 accounting. Thanks
Subject: (usr-tc) PPP/Slip Script problems
From: Richard Roy <rjroy@nbnet.nb.ca>
Date: 1998-02-16 12:17:46
We have a mix of Netserver and hiperarc modems shelve. Here the result I have from dialing to those shelves (after login and password): PPP session on Netserver card - PPP session from (207.179.147.5) to 207.179.147.170 beginning....~= }#=C0!}!}!} =20 PPP session on HiperArc card - ~ }#=C0!}!}!} % =20 SLIP session on Netserver card - SL/IP session from (198.164.98.5) to 198.164.98.168 beginning.... =20 SLIP session on HiperArc card - SLIP connection starting. Your IP address 207.179.158.16. For customers using trumpet winsock, there are looking for the word "to". Works for netserver only. One of the problem is telling the customer to change his/her script. I wonder why 3Com changed it... Richard Roy (rjroy@nbnet.nb.ca) NBTel Internet Technical Analyst / Analyste Technique
Subject: (usr-tc) double up program
From: matthew <matthew@the-spa.com>
Date: 1998-02-16 13:26:04
the pricing for the double up cards was schedule to expire last week. has anyone heard if the new price is higher/lower/same? matthew
Subject: Re: (usr-tc) V.90
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-16 14:13:47
Thus spake Terry Kennedy >I posted this question a week or so ago and got a >response but am still unclear on a certain point. Will >the new standard (v.90) work with unmodified versions >of X2 modems? KFlex modems? It should work (unless USR has lost all concept of backwards compatibility) with x2. I would be extremely surprised to find that it works with kflex...though they might implement that down the road (similar to the v.everything solution they did with 28.8). Worst case, the modems will fall back to v.34 speeds, but I would be extremely surprised if they broke x2 backward compatibility. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Changing radius ports?
From: Support <support@drfast.net>
Date: 1998-02-16 14:16:27
Thanks! Yes I ment log group selection group2
Subject: (usr-tc) New Bundles!!
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-16 14:30:43
Here's some news we can all enjoy!! Some great new bundles from USR, now that the "double up" program is over - we have NEW double up's, and a new replacement for the old 2059 bundles(thank goodness!) Not sure how available any of this stuff is yet, but WOW! 80-002861-00 Entry Level HiPer Bundle with ARC Card - 48 ports* 2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface HiPer Access Router Card with Dual 10/100 Ethernet NIC chassis with integrated fan try, 1-70amp PSU and NMC. Supports IP protocol. This promotional bundle is only available for sale at this price between 2/16/98 and 4/30/98. Only five (5) per customer. This bundle qualifies for the standard warranty. No service or installation is included. Service may be purchased as a separate line item. PRICE: $10,800.00 80-002457-00 Double Up with HiPer Access Router Card - 48 ports 2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface, including HiPer Access Router Card with Dual 10/100 Ethernet NIC and NMC Memory Upgrade kit. Supports IP protocol. 80-002457-00 Double Up with HiPer Access Router Card - 48 ports $22,000.00 ============================================================================ | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: Re[2]: (usr-tc) V.90
From: Terry Kennedy <terry@olypen.com>
Date: 1998-02-16 14:57:39
Great exactly what I needed, thanks. -----Original Message----- > OK, let's see if I can answer all of the questions. (More info is on > our web site) > > 1)Sportsters will ship before the Total Control Upgrade. This is > simply do to less complex regression testing, etc. > > 2)All x2 customers (sportster, TC, etc.) are entitled to a FREE > upgrade to v.90. > > 3)All 3Com products will continue to support both x2 and v.90. > > 4)New sportster modems will support both protocols, so if ISP's are > not running v.90 yet, customers using 3Com modems can fall back to x2. > > 5)3Com products will not support k56flex > > 6)Lastly, ask other modem and NAS vendors to put their position in > writing. Ask them if . . . > > 1)They will provide free upgrades > 2)They will guarantee software upgrades for the NAS > 3)They will guarantee software upgrades that include k56flex and > v.90 running in the same software load at the same time! > > I hope this helps > > Pat > > >______________________________ Reply Separator _________________________________ >Subject: Re: (usr-tc) V.90 >Author: Jeff Mcadams <jeffm@iglou.com> at Internet >Date: 2/16/98 2:13 PM > > >Thus spake Terry Kennedy >>I posted this question a week or so ago and got a >>response but am still unclear on a certain point. Will >>the new standard (v.90) work with unmodified versions >>of X2 modems? KFlex modems? > >It should work (unless USR has lost all concept of backwards >compatibility) with x2. I would be extremely surprised to find that it >works with kflex...though they might implement that down the road >(similar to the v.everything solution they did with 28.8). > >Worst case, the modems will fall back to v.34 speeds, but I would be >extremely surprised if they broke x2 backward compatibility. >-- >Jeff McAdams Email: jeffm@iglou.com >Chief Network Administrator Voice: (502) 966-3848 >IgLou Internet Services (800) 436-4456 > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.
Subject: Re[2]: (usr-tc) V.90
From: phenkle@usr.com
Date: 1998-02-16 16:01:57
--IMA.Boundary.923866788 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part OK, let's see if I can answer all of the questions. (More info is on our web site) 1)Sportsters will ship before the Total Control Upgrade. This is simply do to less complex regression testing, etc. 2)All x2 customers (sportster, TC, etc.) are entitled to a FREE upgrade to v.90. 3)All 3Com products will continue to support both x2 and v.90. 4)New sportster modems will support both protocols, so if ISP's are not running v.90 yet, customers using 3Com modems can fall back to x2. 5)3Com products will not support k56flex 6)Lastly, ask other modem and NAS vendors to put their position in writing. Ask them if . . . 1)They will provide free upgrades 2)They will guarantee software upgrades for the NAS 3)They will guarantee software upgrades that include k56flex and v.90 running in the same software load at the same time! I hope this helps Pat ______________________________ Reply Separator _________________________________ Author: Jeff Mcadams <jeffm@iglou.com> at Internet Thus spake Terry Kennedy >I posted this question a week or so ago and got a >response but am still unclear on a certain point. Will >the new standard (v.90) work with unmodified versions >of X2 modems? KFlex modems? It should work (unless USR has lost all concept of backwards compatibility) with x2. I would be extremely surprised to find that it works with kflex...though they might implement that down the road (similar to the v.everything solution they did with 28.8). Worst case, the modems will fall back to v.34 speeds, but I would be extremely surprised if they broke x2 backward compatibility. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456 - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. --IMA.Boundary.923866788 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4E891CA0; Mon, 16 Feb 98 13:21:46 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id NAA27438; Mon, 16 Feb 1998 13:02:35 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0y4VzG-0003WT-00; Mon, 16 Feb 1998 12:13:54 -0700 Received: from iglou.com [192.107.41.3] (exim) by lists.xmission.com with smtp (Exim 1.73 #4) id 0y4VzB-0003VW-00; Mon, 16 Feb 1998 12:13:49 -0700 Received: from jeffm by iglou.com with local (8.7.3/8.6.12) id 0y4Vz9-0002al-00; Mon, 16 Feb 1998 14:13:47 -0500 In-Reply-To: <m0y4VuR-0005KOC@mail.olypen.com> from "Terry Kennedy" at Feb 16, 98 11:28:40 am X-Mailer: ELM [version 2.4 PL25] Content-Type: text Message-Id: <E0y4Vz9-0002al-00@iglou.com> Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.923866788--
Subject: Re: (usr-tc) New Bundles!!
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-16 16:07:14
The first one is a "You can only buy 5 of them" special bundle. The 2nd one looks like it is a standard part number, that will be discountable by resellers, and you can buy as many as you want to. > yeah we just got this info and were confused... > > the first one for $10,800 includes 2 hiper dsp cards, 1 hiper arc card, 1 > nmc card and a chassis w/70am power supply. > > the second one for $22,000 is just 2 hiper dsp cards, 1 hiper arc card and > the nmc memory upgrades. > > the second one gives you less for $10k more...??? > > matthew > > At 02:30 PM 2/16/98 -0600, you wrote: > >Here's some news we can all enjoy!! Some great new bundles from USR, now > >that the "double up" program is over - we have NEW double up's, and a new > >replacement for the old 2059 bundles(thank goodness!) > > > >Not sure how available any of this stuff is yet, but WOW! > > > >---------------------------------------------------------------------------- > >80-002861-00 Entry Level HiPer Bundle with ARC Card - 48 ports* > >2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface > >HiPer Access Router Card with Dual 10/100 Ethernet NIC > >chassis with integrated fan try, 1-70amp PSU and NMC. Supports IP > protocol. > > > >This promotional bundle is only available for sale at this price > > >between 2/16/98 and 4/30/98. Only five (5) per customer. This bundle > >qualifies for the standard warranty. No service or installation is > >included. Service may be purchased as a separate line item. > > > >PRICE: $10,800.00 > >---------------------------------------------------------------------------- > > > > >80-002457-00 Double Up with HiPer Access Router Card - 48 ports > >2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface, > including HiPer > >Access Router Card with Dual 10/100 Ethernet NIC and NMC Memory Upgrade > > >kit. Supports IP protocol. > > > > > >80-002457-00 Double Up with HiPer Access Router Card - 48 ports > >$22,000.00 > >=========================================================================== > = > > > > > > > > >-------------------------------------------------------------------------- > >| Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > >| Executive Vice President - Exec-PC, Inc. | > >-------------------------------------------------------------------------- > > > > > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: Re[2]: (usr-tc) V.90
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-16 16:30:05
How about one more question: when will the upgrade be available? On Mon, 16 Feb 1998 phenkle@usr.com wrote: > OK, let's see if I can answer all of the questions. (More info is on > our web site) > > 1)Sportsters will ship before the Total Control Upgrade. This is > simply do to less complex regression testing, etc. > > 2)All x2 customers (sportster, TC, etc.) are entitled to a FREE > upgrade to v.90. > > 3)All 3Com products will continue to support both x2 and v.90. > > 4)New sportster modems will support both protocols, so if ISP's are > not running v.90 yet, customers using 3Com modems can fall back to x2. > > 5)3Com products will not support k56flex > > 6)Lastly, ask other modem and NAS vendors to put their position in > writing. Ask them if . . . > > 1)They will provide free upgrades > 2)They will guarantee software upgrades for the NAS > 3)They will guarantee software upgrades that include k56flex and > v.90 running in the same software load at the same time! > > I hope this helps > > Pat > > > ______________________________ Reply Separator _________________________________ > Subject: Re: (usr-tc) V.90 > Author: Jeff Mcadams <jeffm@iglou.com> at Internet > Date: 2/16/98 2:13 PM > > > Thus spake Terry Kennedy > >I posted this question a week or so ago and got a > >response but am still unclear on a certain point. Will > >the new standard (v.90) work with unmodified versions > >of X2 modems? KFlex modems? > > It should work (unless USR has lost all concept of backwards > compatibility) with x2. I would be extremely surprised to find that it > works with kflex...though they might implement that down the road > (similar to the v.everything solution they did with 28.8). > > Worst case, the modems will fall back to v.34 speeds, but I would be > extremely surprised if they broke x2 backward compatibility. > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) New Bundles!!
From: matthew <matthew@the-spa.com>
Date: 1998-02-16 16:37:57
yeah we just got this info and were confused... the first one for $10,800 includes 2 hiper dsp cards, 1 hiper arc card, 1 nmc card and a chassis w/70am power supply. the second one for $22,000 is just 2 hiper dsp cards, 1 hiper arc card and the nmc memory upgrades. the second one gives you less for $10k more...??? matthew At 02:30 PM 2/16/98 -0600, you wrote: >Here's some news we can all enjoy!! Some great new bundles from USR, now >that the "double up" program is over - we have NEW double up's, and a new >replacement for the old 2059 bundles(thank goodness!) > >Not sure how available any of this stuff is yet, but WOW! > >---------------------------------------------------------------------------- >80-002861-00 Entry Level HiPer Bundle with ARC Card - 48 ports* >2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface >HiPer Access Router Card with Dual 10/100 Ethernet NIC >chassis with integrated fan try, 1-70amp PSU and NMC. Supports IP protocol. > >This promotional bundle is only available for sale at this price >between 2/16/98 and 4/30/98. Only five (5) per customer. This bundle >qualifies for the standard warranty. No service or installation is >included. Service may be purchased as a separate line item. > >PRICE: $10,800.00 >---------------------------------------------------------------------------- > >80-002457-00 Double Up with HiPer Access Router Card - 48 ports >2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface, including HiPer >Access Router Card with Dual 10/100 Ethernet NIC and NMC Memory Upgrade >kit. Supports IP protocol. > >80-002457-00 Double Up with HiPer Access Router Card - 48 ports >$22,000.00 >=========================================================================== = > > > >-------------------------------------------------------------------------- >| Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | >| Executive Vice President - Exec-PC, Inc. | >-------------------------------------------------------------------------- > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) New Bundles!!
From: matthew <matthew@the-spa.com>
Date: 1998-02-16 17:16:00
true but still a big difference in price. i ordered a bunch of the first one as soon as i saw the pricing. matthew At 04:07 PM 2/16/98 -0600, you wrote: >The first one is a "You can only buy 5 of them" special bundle. >The 2nd one looks like it is a standard part number, that will be >discountable by resellers, and you can buy as many as you want to. > >---------------------------------------------------------------------- > > >> yeah we just got this info and were confused... >> >> the first one for $10,800 includes 2 hiper dsp cards, 1 hiper arc card, 1 >> nmc card and a chassis w/70am power supply. >> >> the second one for $22,000 is just 2 hiper dsp cards, 1 hiper arc card and >> the nmc memory upgrades. >> >> the second one gives you less for $10k more...??? >> >> matthew >> >> At 02:30 PM 2/16/98 -0600, you wrote: >> >Here's some news we can all enjoy!! Some great new bundles from USR, now >> >that the "double up" program is over - we have NEW double up's, and a new >> >replacement for the old 2059 bundles(thank goodness!) >> > >> >Not sure how available any of this stuff is yet, but WOW! >> > >> >---------------------------------------------------------------------------- >> >80-002861-00 Entry Level HiPer Bundle with ARC Card - 48 ports* >> >2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface >> >HiPer Access Router Card with Dual 10/100 Ethernet NIC >> >chassis with integrated fan try, 1-70amp PSU and NMC. Supports IP >> protocol. >> > >> >This promotional bundle is only available for sale at this price >> >> >between 2/16/98 and 4/30/98. Only five (5) per customer. This bundle >> >qualifies for the standard warranty. No service or installation is >> >included. Service may be purchased as a separate line item. >> > >> >PRICE: $10,800.00 >> >---------------------------------------------------------------------------- >> > >> >> >80-002457-00 Double Up with HiPer Access Router Card - 48 ports >> >2-HiPer DSP modem cards (24 ports each) with T1/PRI NIC interface, >> including HiPer >> >Access Router Card with Dual 10/100 Ethernet NIC and NMC Memory Upgrade >> >> >kit. Supports IP protocol. >> >> > >> >> >80-002457-00 Double Up with HiPer Access Router Card - 48 ports >> >$22,000.00 >> >=========================================================================== >> = >> >> > >> > >> > >> >-------------------------------------------------------------------------- >> >| Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | >> >| Executive Vice President - Exec-PC, Inc. | >> >-------------------------------------------------------------------------- >> > >> > >> > >> >- >> > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> > with "unsubscribe usr-tc" in the body of the message. >> > For information on digests or retrieving files and old messages send >> > "help" to the same address. Do not use quotes in your message. >> > >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the message. >> For information on digests or retrieving files and old messages send >> "help" to the same address. Do not use quotes in your message. >> > >-------------------------------------------------------------------------- >| Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | >| Executive Vice President - Exec-PC, Inc. | >-------------------------------------------------------------------------- > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) HiPer Modem Reboot
From: Marshall Morgan <marshall@netdoor.com>
Date: 1998-02-16 18:22:11
------ =_NextPart_000_01BD3B07.CCCD7A30 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Using v1.0.7 of the HDM modem code in a double up scenario. Using CT1 configuration with a modified generic profile as follows: Framing Mode dsx1ESF Line Coding Options dsx1B8ZS Send Code dsx1SendNoCode Circuit Identifier Loopback Configuration dsx1NoLoop Signal Mode robbedBit Transmit Clock Source loopTiming NIC Type longHaul Response to Remote Loopback ignore Jitter Attenuation attenJitterOnTxmtr Transmit Line Build Out dB0pt0 Dial In Address noAddress Dial In/Out Trunk Start Signal Type wink Ack Wink On Dial In Address Info Received disabled Dial Out Address Delay 70 Dial In/Out Trunk Type eAndMTypeII Primary Switch Type priSw5ESS Idle Byte Pattern 254 Receiver Gain dB26 Tone Type dtmf Number of DTMF Tones 4 Inventoy as well: We just started using these HDM modems today and the second of two rebooted for some reason. Anyone got any ideas? We are co-located with a telco and the switch is only a few feet from us. All other USR TC's are using TCS 3.01 with i386 cards and no HiPer cards without issue. Marshall Morgan http://www.netdoor.com 601.969.1434 Ext. #28 | Fax 601.969.3838 | 800.952.1570 Ext. #28 ------ =_NextPart_000_01BD3B07.CCCD7A30--
Subject: Re[4]: (usr-tc) V.90
From: phenkle@usr.com
Date: 1998-02-16 20:03:33
--IMA.Boundary.802186788 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Alpha trials with quad modems have been underway. Official Beta began today. HiPer DSP Beta is expected in about 10 days. Release of the software upgrade is expected in March. That is as close as I can get for now. Pat ______________________________ Reply Separator _________________________________ Author: Lee Kuo <lee@cosmo.mitec.net> at Internet How about one more question: when will the upgrade be available? On Mon, 16 Feb 1998 phenkle@usr.com wrote: > OK, let's see if I can answer all of the questions. (More info is on > our web site) > > 1)Sportsters will ship before the Total Control Upgrade. This is > simply do to less complex regression testing, etc. > > 2)All x2 customers (sportster, TC, etc.) are entitled to a FREE > upgrade to v.90. > > 3)All 3Com products will continue to support both x2 and v.90. > > 4)New sportster modems will support both protocols, so if ISP's are > not running v.90 yet, customers using 3Com modems can fall back to x2. > > 5)3Com products will not support k56flex > > 6)Lastly, ask other modem and NAS vendors to put their position in > writing. Ask them if . . . > > 1)They will provide free upgrades > 2)They will guarantee software upgrades for the NAS > 3)They will guarantee software upgrades that include k56flex and > v.90 running in the same software load at the same time! > > I hope this helps > > Pat > > > ______________________________ Reply Separator _________________________________ > Subject: Re: (usr-tc) V.90 > Author: Jeff Mcadams <jeffm@iglou.com> at Internet > Date: 2/16/98 2:13 PM > > > Thus spake Terry Kennedy > >I posted this question a week or so ago and got a > >response but am still unclear on a certain point. Will > >the new standard (v.90) work with unmodified versions > >of X2 modems? KFlex modems? > > It should work (unless USR has lost all concept of backwards > compatibility) with x2. I would be extremely surprised to find that it > works with kflex...though they might implement that down the road > (similar to the v.everything solution they did with 28.8). > > Worst case, the modems will fall back to v.34 speeds, but I would be > extremely surprised if they broke x2 backward compatibility. > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. --IMA.Boundary.802186788 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4E8C0C50; Mon, 16 Feb 98 16:42:13 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id QAA06495; Mon, 16 Feb 1998 16:22:57 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0y4Z9F-0002fu-00; Mon, 16 Feb 1998 15:36:25 -0700 Received: from mitec.net [208.130.151.6] (qmailr) by lists.xmission.com with smtp (Exim 1.73 #4) id 0y4Z9C-0002ep-00; Mon, 16 Feb 1998 15:36:23 -0700 Received: (qmail 16299 invoked by uid 505); 16 Feb 1998 22:30:05 -0000 In-Reply-To: <4E8BE690.3000@usr.com> Message-ID: <Pine.LNX.3.95.980216162909.16243A-100000@cosmo.mitec.net> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.802186788--
Subject: Re: (usr-tc) HiPer Modem Reboot
From: Brian <signal@shreve.net>
Date: 1998-02-16 21:35:04
we saw this problem with release hyperarc code, and are now running ER code (4.0.80) which seems much more stable. Brian On Mon, 16 Feb 1998, Marshall Morgan wrote: > Using v1.0.7 of the HDM modem code in a double up scenario. Using CT1 > configuration with a modified generic profile as follows: > > Framing Mode dsx1ESF > Line Coding Options dsx1B8ZS > Send Code dsx1SendNoCode > Circuit Identifier > Loopback Configuration dsx1NoLoop > Signal Mode robbedBit > Transmit Clock Source loopTiming > NIC Type longHaul > Response to Remote Loopback ignore > Jitter Attenuation attenJitterOnTxmtr > Transmit Line Build Out dB0pt0 > Dial In Address noAddress > Dial In/Out Trunk Start Signal Type wink > Ack Wink On Dial In Address Info Received disabled > Dial Out Address Delay 70 > Dial In/Out Trunk Type eAndMTypeII > Primary Switch Type priSw5ESS > Idle Byte Pattern 254 > Receiver Gain dB26 > Tone Type dtmf > Number of DTMF Tones 4 > > Inventoy as well: > > > > We just started using these HDM modems today and the second of two rebooted for > some reason. Anyone got any ideas? We are co-located with a telco and the > switch is only a few feet from us. All other USR TC's are using TCS 3.01 with > i386 cards and no HiPer cards without issue. > > Marshall Morgan > > http://www.netdoor.com > 601.969.1434 Ext. #28 | Fax 601.969.3838 | 800.952.1570 Ext. #28 > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) HELP! USR Total Control Rackmount32 Question
From: beccom@aol.com
Date: 1998-02-17 02:31:47
I have a total control rackmount32 and it has not been used for a while by another dept. We have acquired the chassis and would like to find out the maximum speed of these modems. The chassis holds 16 dual analog cards with separate RS232 nics in the back of the chassis. The NICS at the back also have analog ports. The chassis comes complete with RCU, dual Power supply and a CIU card. To test it I connected it to an NT 4.0 server that was running RAS. After trying several modems, the one that worked the best was a Courier v.32 w/ASL but I could only connect at 12,000 everytime not higher no lower. What I am thinking is that it has some sort of initialization string that is configured in the CIU and it is set for no higher than 12,000. I was reading the manual and it says that I can control the Chassis by connecting the DB9 to RJ45 serial connector that came included with the CIU....but I was not aware of that and have to go look for it. When I entered the ATI7 command it showed that it had a 16 MHz chip in it. and a few of the dates were from 93 and 94. I am hoping that it is at least a 28,800 modem so we can use this box. Can you please let me know if there is a command that I can use to check the speed or do I have to use the rj45 serial cable that was supplied to change initializations strings? I have worked on this for a couple of days now and I am running out of options. Any support that you can supply me with would be greatly appreciated. If you need to contact me please email BECCOM@aol.com Thanks Again Chris B
Subject: Re: (usr-tc) 3.7.24 reboots?
From: Brian <signal@shreve.net>
Date: 1998-02-17 08:14:24
On Tue, 17 Feb 1998, Bob Purdon wrote: > > Has anyone seen spontaneous reboots with 3.7.24? > > I've observed (last night) our only 3.7.24 rack spontaneously reboot. > It's not done it since. With 3.5.33 this chassis ran for months, only to > go down for a flash upgrade. > > I'm hoping this isn't a common problem, and that it was only the NETserver > getting crabby because I upgraded it's Quad cards without asking it really > nicely first :-) no problems so far. > > Regards, > > Bob Purdon, > Technical Manager, > Southern Internet Services. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) 3.7.24 reboots?
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-17 09:24:52
> Has anyone seen spontaneous reboots with 3.7.24? > > I've observed (last night) our only 3.7.24 rack spontaneously reboot. > It's not done it since. With 3.5.33 this chassis ran for months, only to > go down for a flash upgrade. > > I'm hoping this isn't a common problem, and that it was only the NETserver > getting crabby because I upgraded it's Quad cards without asking it really > nicely first :-) While I haven't expirenced any reboots, I will say that I ALWAYS reboot my netserver after flashing the modems. The netserver just doesn't seem to come back in a "normal" state unless you do that. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) V.90
From: jason_kelton@3com.com
Date: 1998-02-17 09:28:15
As I have been told, the TCENH will have the capcity to accept both V.90 and x2 calls. I don't know about the client side. Regards, Jason. terry@olypen.com on 17/02/98 05:28:40 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) I posted this question a week or so ago and got a response but am still unclear on a certain point. Will the new standard (v.90) work with unmodified versions of X2 modems? KFlex modems? In other works, when the code becomes available and we load it on the netservers, will the customers be forced to upgrade their modems to v.90 before they get a 56k connection again? Although it's conceivable that most customers would upgrade eventually, it would be a nightmare to load the code one night and anwer the phones the next day and start explaining to them that they need to upgrade even if it is just a software solution. Not to mention the ones that need a hardware upgrade. Any one care to shed some light on this. There is nothing at totalservice.com that helps to explain this. Terry Kennedy OlyPen, Inc. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) V.90
From: jason_kelton@3com.com
Date: 1998-02-17 09:28:15
As I have been told, the TCENH will have the capcity to accept both V.90 and x2 calls. I don't know about the client side. Regards, Jason. terry@olypen.com on 17/02/98 05:28:40 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) I posted this question a week or so ago and got a response but am still unclear on a certain point. Will the new standard (v.90) work with unmodified versions of X2 modems? KFlex modems? In other works, when the code becomes available and we load it on the netservers, will the customers be forced to upgrade their modems to v.90 before they get a 56k connection again? Although it's conceivable that most customers would upgrade eventually, it would be a nightmare to load the code one night and anwer the phones the next day and start explaining to them that they need to upgrade even if it is just a software solution. Not to mention the ones that need a hardware upgrade. Any one care to shed some light on this. There is nothing at totalservice.com that helps to explain this. Terry Kennedy OlyPen, Inc. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) power requirement?
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-17 09:29:39
> I want to use fully loaded hiperdsp & hiperarc. i.e. 14DSP > and 2HiperARC with 48V DC. From website, I have to use 130Amp DC > Does this mean I have to have at least 130Amp power source? > I don't think this sytem use that kind of amount amps. > What they really mean? And is it possible to have two os on hiperarc? > If possible I want to swich back to old os if new code is something > wrong. > Any help is much apprecaited. The 130A rating is based on the 5v power it produces. So, approx 13.5A would be the needs at 48v, and I doubt it will ever really get that high. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) 3.7.24 reboots?
From: Chris Getman <getman@frontiernet.net>
Date: 1998-02-17 10:53:12
I haven't seen any yet. Here I have noticed that I will loose one to two racks a week on the older code. The racks just drop off the network. Either a power cycle or a hardware reset from the TCM gets them going again. I never loose any configurations. Weird. I am hoping that the new code helps with that problem. -Chris > > Has anyone seen spontaneous reboots with 3.7.24? > > > > I've observed (last night) our only 3.7.24 rack spontaneously reboot. > > It's not done it since. With 3.5.33 this chassis ran for months, only to > > go down for a flash upgrade. > > > > I'm hoping this isn't a common problem, and that it was only the NETserver > > getting crabby because I upgraded it's Quad cards without asking it really > > nicely first :-) > > While I haven't expirenced any reboots, I will say that I ALWAYS reboot my > netserver after flashing the modems. The netserver just doesn't seem to > come back in a "normal" state unless you do that. >
Subject: (usr-tc) no message on ARC engineering 4.0.80
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-17 10:54:13
It seems that after a couple days of running on 4.0.80 engineering release, the ARC stops sending any of the "message" to the dialin callers, and it also doesn't echo back anything you type. You can still login if you type the information anyway. Looks like I have to back out to 4.0.89. -- Pete XMission
Subject: Re: (usr-tc) no message on ARC engineering 4.0.80
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-17 11:40:25
Charles Hill said once upon a time: > >Pete, > >I have seen this behavior once in 4.0.95, so I'm not sure 4.0.89 will be a >fix. > >-CH > >On Tue, 17 Feb 1998, Pete Ashdown wrote: > >> It seems that after a couple days of running on 4.0.80 engineering release, >> the ARC stops sending any of the "message" to the dialin callers, and it >> also doesn't echo back anything you type. You can still login if you type >> the information anyway. Looks like I have to back out to 4.0.89. On my rack running 4.0.89, I never see this problem. Of course, what I do see are random crashes, which is what 4.0.80 was supposed to fix.
Subject: Re: (usr-tc) 3.7.24 reboots?
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-17 11:42:44
On Tue, 17 Feb 1998, Chris Getman wrote: > I haven't seen any yet. Here I have noticed that I will loose one to > two racks a week on the older code. The racks just drop off the network. > Either a power cycle or a hardware reset from the TCM gets them going > again. I never loose any configurations. Weird. I am hoping that the > new code helps with that problem. > > -Chris > > We fixed that problem here by sending all our netservers with FC3 chips on them in for RMA. When you get them back they have FC2 chips on them instead. USR told us it was a hardware bug and we have not had the problem since. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) no message on ARC engineering 4.0.80
From: Charles Hill <chill@ionet.net>
Date: 1998-02-17 12:22:25
Pete, I have seen this behavior once in 4.0.95, so I'm not sure 4.0.89 will be a fix. -CH On Tue, 17 Feb 1998, Pete Ashdown wrote: > It seems that after a couple days of running on 4.0.80 engineering release, > the ARC stops sending any of the "message" to the dialin callers, and it > also doesn't echo back anything you type. You can still login if you type > the information anyway. Looks like I have to back out to 4.0.89. > > -- > Pete > XMission > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) MPIP on 3.7.24 and ISDN-Modem
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-17 16:05:37
Practically nobody can use MPIP since we upgraded to 3.7.24 and put ISDN processing on the quads. I'm going to reboot everything tomorrow to see if that helps, but I'm wondering if I'm missing something in regards to MPIP and ISDN-Modem processing.
Subject: Re: (usr-tc) telnet wierdness
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-17 17:03:15
On Mon, 16 Feb 1998, Bob Purdon wrote: > > > Did you try this 100 times in sequence (i.e, connecting with some > > utility, disconnecting, and then connecting again?) > > OK, I've just put 3.7.24 on one of our chassis and the behaviour has > changed. Previously at 3.6.28 and 3.5.33, when using RIPv1 routing, I had > a 1 second delay on connecting. Changing to RIPv2 made connections > instantaneous. > > At 3.7.24 it's instantaneous regardless :-) > Even with "set ripv2 on" set, I have yet to see the telnet delay disappear completely. I still get it on every initial telnet connection. Command> version U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.7.24 Build date: Dec 31 1997 Build time: 13:12:45 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Standard Perhaps I need to reboot the netserver with ripv2 enabled, there is another command that needs to be set, or its hardware revision related? - lv
Subject: (usr-tc) HDM woes (again)
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-17 18:37:37
A couple of weeks ago I sent a message to the list re: inability to get ISDN connections to HDM cards. Since then I have worked with our telco (getting test numbers, etc) and played with numerous configuration settings. I have most definitely come to the conclusion that it IS a config problem. I can make the same circuits that DON'T work on HDM cards accept ISDN w/ no problem on a dual t1/pri. Here's my ISDN related HDM configuration: (using template 1 for all channels) V110 Rate Adaption (S67.0) : disable Force Fixed Network Rate (S67.1): notForced Force Network Rate Speed : kbps64 Enable 45-64 Second Link Delay : noDelay Analog Calls Over Digital : allow Async PPP/Sync PPP Conversion : allow X.75 (S68.5) : disable Set Data Mode of Modem (*V2=x) : autodetect Set Originate HDLC Protocol : v120 Set Originate Non-HDLC Protocol : none Set Originate Analog Modem/Fax : analogModemFax V120 : disable X75 Frame Size : 2048 X75 Window Size : 2 I'm not blocking any calls, and analog connections are no problem on the HDM. Anyone have any ideas? Thanks ... Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) MPIP on 3.7.24 and ISDN-Modem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-17 18:44:30
Thus spake Pete Ashdown >Practically nobody can use MPIP since we upgraded to 3.7.24 and put ISDN >processing on the quads. I'm going to reboot everything tomorrow to see if >that helps, but I'm wondering if I'm missing something in regards to MPIP >and ISDN-Modem processing. Man, 3.7.24 is the first time MPIP has worked correctly for us...we're running ISDN on the Munich card though....not sure how much that would affect it. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) power requirement?
From: MyungSik Kim <daesung@soback.kornet.nm.kr>
Date: 1998-02-17 20:28:06
I want to use fully loaded hiperdsp & hiperarc. i.e. 14DSP and 2HiperARC with 48V DC. From website, I have to use 130Amp DC Does this mean I have to have at least 130Amp power source? I don't think this sytem use that kind of amount amps. What they really mean? And is it possible to have two os on hiperarc? If possible I want to swich back to old os if new code is something wrong. Any help is much apprecaited.
Subject: (usr-tc) psu trap mean?
From: MyungSik Kim <daesung@soback.kornet.nm.kr>
Date: 1998-02-17 20:43:47
I set PSU failed event to set trap to alarm server and tried to removed PSU and pull out the power code. Then time I couldn't see any alram trap. Only the out of voltage or PSU is broken then time I can see trap message?
Subject: Re: (usr-tc) [Q] Secondary IP address on net0 [Total Control Chassis]?
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-17 20:58:56
No... krish On Tue, 17 Feb 1998, DWX Network Operations wrote: > Greetings. > > Excuse what is no doubt a simple question. > > Does anyone know an/the way to assign a secondary ip address to the net0 > port of a NetServer Card? > > On Cisco boxes, this is simple with a command such as: > > ip address 192.68.123.242 255.255.255.0 secondary > > > Does there exist such a beast for the NetServer Card? > > Thanks. > Tracy Hinshaw, > DWX Internet > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) [Q] Secondary IP address on net0 [Total Control Chassis]?
From: DWX Network Operations <tracy@dwx.com>
Date: 1998-02-17 21:05:09
Greetings. Excuse what is no doubt a simple question. Does anyone know an/the way to assign a secondary ip address to the net0 port of a NetServer Card? On Cisco boxes, this is simple with a command such as: ip address 192.68.123.242 255.255.255.0 secondary Does there exist such a beast for the NetServer Card? Thanks. Tracy Hinshaw, DWX Internet
Subject: (usr-tc) 3.7.24 reboots?
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-17 21:21:11
Has anyone seen spontaneous reboots with 3.7.24? I've observed (last night) our only 3.7.24 rack spontaneously reboot. It's not done it since. With 3.5.33 this chassis ran for months, only to go down for a flash upgrade. I'm hoping this isn't a common problem, and that it was only the NETserver getting crabby because I upgraded it's Quad cards without asking it really nicely first :-) Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: Re: (usr-tc) [Q] Secondary IP address on net0 [Total Control Chassis]?
From: Charles Hill <chill@ionet.net>
Date: 1998-02-17 21:38:29
Version 4.x allows multiple addresses per interface. I don't believe it's possible in 3.x. -CH On Tue, 17 Feb 1998, DWX Network Operations wrote: > Greetings. > > Excuse what is no doubt a simple question. > > Does anyone know an/the way to assign a secondary ip address to the net0 > port of a NetServer Card? > > On Cisco boxes, this is simple with a command such as: > > ip address 192.68.123.242 255.255.255.0 secondary > > > Does there exist such a beast for the NetServer Card? > > Thanks. > Tracy Hinshaw, > DWX Internet > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) Enhanced routing
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-17 22:26:22
I read a message here a few weeks ago about consolidating RIP routes on a chassis into 1 large route. Just add an IP pool, add a netmask entry, turn enh_routing on and voila! It works great, BUT: Anyone with an ip outside that pool no longer has their ip address broadcasted via RIP. This is a problem. I really like the enhanced routing, but does anyone know how to make it keep broadcasting the ip's that are not in the pool? example: ip pool entry: 206.175.52.1 size 61 netmask entry: 206.175.52.0 255.255.255.192 other ip: 205.156.197.150 255.255.255.255 (this route doesn't get broadcasted) Some of these ip's still get routed properly, but the route never shows up on our Cisco. Thanks for any ideas, or more info on what exactly the enhanced routing does (since it doesn't appear to be very well documented). Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com
Subject: Re: (usr-tc) MPIP on 3.7.24 and ISDN-Modem
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-17 22:52:26
At 04:05 PM 2/17/98 -0700, Pete Ashdown wrote: >Practically nobody can use MPIP since we upgraded to 3.7.24 and put ISDN >processing on the quads. I'm going to reboot everything tomorrow to see if >that helps, but I'm wondering if I'm missing something in regards to MPIP >and ISDN-Modem processing. > We use 3.7.24, MPIP, and terminate isdn calls on the quads.. It works as well as it ever did which was never perfect. And of course, not at all the the arcs in our network.. :( allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Enhanced routing
From: Michael Mittelstadt <meek@execpc.com>
Date: 1998-02-17 23:02:35
[Quoth Peter D. Mayer] ] I read a message here a few weeks ago about consolidating RIP routes on a ] chassis into 1 large route. Just add an IP pool, add a netmask entry, turn ] enh_routing on and voila! It works great, BUT: Anyone with an ip outside ] that pool no longer has their ip address broadcasted via RIP. This is a ] problem. I really like the enhanced routing, but does anyone know how to ] make it keep broadcasting the ip's that are not in the pool? ] ] example: ] ip pool entry: 206.175.52.1 size 61 ] netmask entry: 206.175.52.0 255.255.255.192 ] other ip: 205.156.197.150 255.255.255.255 (this route doesn't get ] broadcasted) ] Some of these ip's still get routed properly, but the route never shows up ] on our Cisco. I am doing this, and I do see routes for people who are outside of the dialup range. Is the ethernet of your TC in a seperate network from the IP pool? Did you turn on RIPv2 on on the TC? Try turning on RIP debugging on the Cisco, and see if it's just rejecting them for some reason. -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1
Subject: Re: (usr-tc) HELP! USR Total Control Rackmount32 Question
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-17 23:09:22
On Tue, 17 Feb 1998 BECCOM@aol.com wrote: > I have a total control rackmount32 and it has not been used for a while by > another dept. We have acquired the chassis and would like to find out the > maximum speed of these modems. The chassis holds 16 dual analog cards with > separate RS232 nics in the back of the chassis. The NICS at the back also > have analog ports. The chassis comes complete with RCU, dual Power supply and > a CIU card. To test it I connected it to an NT 4.0 server that was running > RAS. After trying several modems, the one that worked the best was a Courier > v.32 w/ASL but I could only connect at 12,000 everytime not higher no lower. > What I am thinking is that it has some sort of initialization string that is > configured in the CIU and it is set for no higher than 12,000. I was reading > the manual and it says that I can control the Chassis by connecting the DB9 to > RJ45 serial connector that came included with the CIU....but I was not aware > of that and have to go look for it. When I entered the ATI7 command it showed > that it had a 16 MHz chip in it. and a few of the dates were from 93 and 94. > I am hoping that it is at least a 28,800 modem so we can use this box. Can > you please let me know if there is a command that I can use to check the speed > or do I have to use the rj45 serial cable that was supplied to change > initializations strings? I have worked on this for a couple of days now and I > am running out of options. Any support that you can supply me with would be > greatly appreciated. If you need to contact me please email BECCOM@aol.com > This rack is a very old rack, and I do not think that we supported 28K - You may have only v.32 and bis modems in the chassi. krish > > Thanks Again > Chris B > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) HDM woes (again)
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-18 00:21:22
At 06:37 PM 2/17/98 -0500, System Administrator wrote: >A couple of weeks ago I sent a message to the list re: inability to get >ISDN connections to HDM cards. Since then I have worked with our >telco (getting test numbers, etc) and played with numerous configuration >settings. I remember, we had your exact same symptom but it was telco related. We could connect via one of the pilot numbers but not the main number. They fixed it quickly. Been connecting fine with isdn ever since.. >I have most definitely come to the conclusion that it IS a config problem. >I can make the same circuits that DON'T work on HDM cards accept ISDN w/ >no problem on a dual t1/pri. heh. the "acid test".. >Here's my ISDN related HDM configuration: Here is ours: (I see two differences) V110 Rate Adaption (S67.0) disable Force Fixed Network Rate (S67.1) notForced Force Network Rate Speed kbps56 <--- Enable 45-65 Second Link Delay noDelay Analog Calls Over Digital (S68.0) allow Anync PPP/Sync PPP Conversion disable <--- X.75 (S68.5) disable Set Data Mode of Modem (*V2=x) autodetect Set Originate HDLC Protocol v120 Set Originate Non-HDLC Protocol none Set Originate Analog Modem/Fax analogModemFax V120 (S68.6) disable X75 Frame Size 2048 X75 Window Size 2 All I know is that it works for us.. We are using HiPer DSP Rev. 1.0.7. HiperArc 4.0.80 and NMC 5.2.2.. Here's our span config.. span1> display ccrcfig Span1 Configured Signal Mode is (sigmode): MESSAGE ORIENTED Span1 Signal Mode Active is: MESSAGE ORIENTED Span1 DNIS Enable is (dnisena): DNIS ADDRESS Span1 Dial In Out Trunk Start (diotrst): WINK Span1 Dial In Address ACK Wink (daackwnk): ACK WINK DISABLED Span1 Dial Out Address Delay (doadrdly): 70 milliseconds Span1 Dial In Out Trunk Type (dtrnktyp): E&M TYPE II Span1 Configured Switch Type is (swtype): 5ESS Span1 Switch Type Active is: 5ESS Span1 Idle Byte is (idlebyte): 0xFE Span1 Ana Calls Blocked Err Code (ancbec): 58 Span1 Digi Calls Blocked Err Code (dcbec): 58 Span1 No IGWS Avail Err Code (noigwsav): 58 Span1 Chan Blocked Err Code (chanblck): 58 Span1 Block Call Type is (blcaltyp): BLOCK NONE Span1 Tone Type is (tonetype): DTMF TONE Span1 Number Of DTMF Tones is (numdtmft): 4 Span1 Dial Out Select Direction (dseldir): DOWN Span1 Dial Out Next Timeslot (dntslot): 24 Let me know if there are other configs you'd like to see. If it really is a config problem, should be easy to spot.. allen _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: (usr-tc) tcs ramblings
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-18 02:13:42
Well except for the lack of MPIP on the arc and OSPF anywhere, I'm a pretty happy camper at the moment. Never mind that I just bought a stripped chassis last week for half the price of that new (wow) bundle mentioned yesterday. An extra $5.8K would have bought an arc, and 2 hdms.. Oh well, I paid nearly double for a new 1706 right before they dropped a year ago.. Guess that's life in the fast lane.. x2 that is.. But we are 100% usr and I'm not too humble (or embarrassed) to write a stupid X-mas song or sing usr's praises (or bitch) when and where its due. And I hope I haven't wasted bandwidth and your time in the process.. I just wanted to stop and reflect on the state of things and maybe be proud of USR/3com for a while. X2 was a big boom for us. I really sorta hate to see the 56K "wars" over so soon. We came into the market late and it gave us a competitive edge in the marketplace. We had a monopoly on 56k in town for the longest time.. And kflex never did sqat! We sold alot of x2 sportsters in this town indirectly for usr! So the benefit was mutual.. The quake lag thing is a real sore spot with us. We are now 50% hdm/arc and 50% quad/netserver.. I get rather inconsistent results, but overall, quake it playable over either configuration. Not the best pings over a netserver, but much better than we were getting. And 96 modems on a netserver was (is) out of the question or quake is unplayable. I'm interested in putting a second arc in our hdm chassis.. Never did get a "Quake lag upgrade kit" from David Moore as promised but I guess I can consider the new bundle as being what I've waited for. Just wish I had some money left.. (don't worry, I have visa and m/c for ispcon :) We ran after the "double-up" kits thinking hdm's were going up.. Pretty stupid of me actually... HDM's overall are a near perfect idea. I don't see how all you big outfits have lived without them all this time. How do you co-locate 10 tons of A/C at the telco??.. David's probably LOL, more like 100 tons.. Or his might be liquid cooled. heh. I just like the form factor and scalability for an outfit the size of ours.. If only they could have fit 24 micro leds on the card instead of the "VU meter"... they make some pretty cool leds nowadays... Thanks to this list, we haven't called USR very often lately. I put a call in last month at 1am and got a call back within the hour so I'm not complaining. However I was struck by a recent email from iomega where they are being sued over phone support. Here's a snippet: >**** iomega quote **** >c. For the two year period beginning on the >Settlement Effective Date, Iomega shall at its >expense commit to limit "hold" times for technical >support regarding Iomega's Zip, Jaz, and Ditto >drives to ten (10) minutes on an average monthly >basis. If the average "hold" time, as calculated >at month's end, exceeds ten (10) minutes because >call volume is higher than projected, then Iomega >shall use its reasonable efforts to reduce the >"hold" times to ten (10) minutes or below within >forty-five (45) days after the end of the month >during which average "hold" times of greater than >ten (10) minutes were first discovered. >**** end quote **** So customers who spent $100-500 are sueing over hold queue times.. <jeez> And I thought I might be sorta harsh on the USR folk at times... Can't imagine getting so upset as to hire an attorney at $200/hr... A 15 year old brought about the suit with daddy's help.. hmmm.. Maybe it was pro-bono... I am sure MPIP will be out for the arc shortly and we might even get OSPF someday.. I still haven't heard word on when the PM4's might be out so I am pleased with the steps/leap's we have taken. Sorry for rambling I just wanted to say that USR is doing pretty good IMO despite the usual struggles with early code revisions, etc.. I know krish and others have spent time getting us this far and we appreciate it. Looks like it will be TCS for the longhaul... allen P.S. And you gotta love those usr modems. They can actually squeeze 48k bandwidth though the old Bellsouth network in our town; an amazing feat when I stop to think about it!
Subject: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-18 07:18:56
It appears that either making changes to the modem settings or writing to NVRAM causes the HiPer DSP to have a weird interaction with the ARC on ER 4.0.80 and 4.0.72. The symptom is that the modem_group message is not displayed and the connection's login information is not echoed back as it is typed. The workaround is to do a software modem reset after you make changes to either the modems or NVRAM. I am guessing the cause from some testing I did this morning. I think I have finally settled on my settings for my DSP modems, so I don't think I'll have to make further changes. If the bug reoccurs without me changing anything, then I'll report back.
Subject: Re: (usr-tc) Netserver user definitions
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-18 07:24:55
On Wed, 18 Mar 1998, Robert von Bismarck wrote: > I have a couple questions about the Netserver 16 >=20 > I want to use a Netserver 16 with 8 ISDN BRI's for my signup users > (using MS IEAK 4) I do not want this machine to access the network. The > setup is done, it works splendidly, but the Netserver doesn't accept > more than one signup connection. I defined the user in the Netserver, as > I do not want to use a Radius server for this (who needs radius for just > one user=A0?=A0?=A0?) > Any pointers to that=A0? I guess when you say MS IEAK - you mean Microsoft Internet Explorer -=20 Check the IP address of the user, if this is a static IP user the you can= =20 have only one connection. krish >=20 > I'm using Netserver + with 4.0.14 software on it and Netserver Manager > Plus... >=20 > Thanks, >=20 > Robert von Bismarck > Petrel Communication S.A. >=20 > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >=20
Subject: Re: (usr-tc) Need Help getting Framed user from Merit Radius
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-18 07:25:54
On Wed, 18 Mar 1998, Josh Richards <Josh Richards wrote: > On 17 Mar 1998, Jaye Mathisen wrote: > > > > > As I recall, MTU is negotiated as the least common denominator, so > > it should be a nonissue. > > > > However, I also recall reading that MTU negotiation takes place as > > part of the PPP setup stuff, so by the time the radius server > > passes the info back, the MTU has already been fixed. > > > > Scripted logins don't have this problem, since they get the info back > > before the start of PPP negotiation. > > > > YMMV, and I could be completely bonkers. > > Nope - That is correct. The MTU specified in RADIUS is useless for PAP > based PPP logins since PPP has already started before the RADIUS server > even comes into the picture. The above statement is correct for Livingston and NETServer - Not for HiPer ARC. krish > > --jr > > ---- > Josh Richards - <jrichard@livingston.com> - [Beta Engineer] > LUCENT Technologies - Remote Access Business Unit > (formerly Livingston Enterprises, Inc.) > http://www.livingston.com/ > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) 3.7.24 reboots?
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-18 07:42:58
> While I haven't expirenced any reboots, I will say that I ALWAYS reboot > my netserver after flashing the modems. The netserver just doesn't seem > to come back in a "normal" state unless you do that. I may have to start doing that. I prefer not to, as I typically upgrade a chassis over a few days as and when the quads become free. That way i don't have to schedule any downtime :-) Regards, Bob Purdon, Technical Manager, Southern Internet Services.
Subject: (usr-tc) Edgerserver ISDN
From: Tim Buchalka <buchalka@newave.net.au>
Date: 1998-02-18 07:49:39
Hi Jason, I am attempting to get ISDN working over our Edgeserver into a modem (Quad). Have spent some time confirming settings, etc (We are not using a NetServer). Any general comments or things to check for us, as the authentication fails (according to the router). However it appears the ISDN call is not even getting to the modem. Any help appreciated. Thanks Tim
Subject: (usr-tc) converting 45amp to PSU to DC.
From: Henry Moats <nc0419@corp.netcom.com>
Date: 1998-02-18 07:58:25
Does anyone know the process of converting a 45amp chassis to DC power? Thanks. ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________|
Subject: Re: (usr-tc) tcs ramblings
From: Charles Hill <chill@ionet.net>
Date: 1998-02-18 08:29:38
On Wed, 18 Feb 1998, Allen Marsalis wrote: > Well except for the lack of MPIP on the arc and OSPF anywhere, > I'm a pretty happy camper at the moment. Never mind that I Ditto. > HDM's overall are a near perfect idea. I don't see how all you > big outfits have lived without them all this time. How do you > co-locate 10 tons of A/C at the telco??.. The idea is to get the telco to provide A/C in the initial colo contract. :) > If only they could have fit 24 micro leds on the card instead of the "VU > meter"... they make some pretty cool leds nowadays... No, what they should have done is put a little NMC-style screen that just shows the number of channels in use! -CH
Subject: Re: (usr-tc) MPIP on 3.7.24 and ISDN-Modem
From: Brian <signal@shreve.net>
Date: 1998-02-18 08:53:01
On Tue, 17 Feb 1998, Jeff Mcadams wrote: > Thus spake Pete Ashdown > >Practically nobody can use MPIP since we upgraded to 3.7.24 and put ISDN > >processing on the quads. I'm going to reboot everything tomorrow to see if > >that helps, but I'm wondering if I'm missing something in regards to MPIP > >and ISDN-Modem processing. > > Man, 3.7.24 is the first time MPIP has worked correctly for us...we're > running ISDN on the Munich card though....not sure how much that would > affect it. fwiw, mpip has almost always worked for us, but we only scale 3 chassis with mpip, so maybe it has to do with the amount of mpip your doing. We are running 3.7.24 with mpip working well. Brian > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Enhanced routing
From: Brian <signal@shreve.net>
Date: 1998-02-18 08:58:13
On Tue, 17 Feb 1998, Peter D. Mayer wrote: > I read a message here a few weeks ago about consolidating RIP routes on a > chassis into 1 large route. Just add an IP pool, add a netmask entry, turn > enh_routing on and voila! It works great, BUT: Anyone with an ip outside > that pool no longer has their ip address broadcasted via RIP. This is a > problem. I really like the enhanced routing, but does anyone know how to > make it keep broadcasting the ip's that are not in the pool? This is not true. > > example: > ip pool entry: 206.175.52.1 size 61 > netmask entry: 206.175.52.0 255.255.255.192 > other ip: 205.156.197.150 255.255.255.255 (this route doesn't get > broadcasted) > Some of these ip's still get routed properly, but the route never shows up > on our Cisco. > > Thanks for any ideas, or more info on what exactly the enhanced routing does > (since it doesn't appear to be very well documented). > Not sure what can be going on but, here is what we have: set net0 routing on # broadcast and listen to RIP set ripv2 on # use RIPv2 set enh_routing on # used enhanced routing add ippool dialup1 208.214.44.1 62 add netmask 208.214.44.0 255.255.255.192 Perhaps a reboot is needed after all this jazz, don't remember, but as you can see users are given IP's out of the IP pool declared above. We also have static IP's assigned (on *any* hub) that come out of 208.214.45.240/28, and those get added to the tables and broadcasted just fine. 1. What network is your net0 address on? According to Michael Mittelstadt <meek@execpc.com> who orignally posted this idea, "The TC and it's dialup pool must not be in overlapping networks..." Brian > Peter D. Mayer > NetWalk Tech Support > dmayer@netwalk.com > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Newbie questions about HiPerARC
From: Brian <signal@shreve.net>
Date: 1998-02-18 09:12:18
On Wed, 18 Feb 1998, Robert von Bismarck wrote: > Couple Newbie questions=A0: >=20 > -=09Is there a way to replace MPIP, or to forbid MLPPP ISDN > connections=A0? You do this in RADIUS via the Port-Limit attribute. > -=09How do I add a user with a fixed IP address that is not in the > standard dial-in pool=A0? You do this in RADIUS also by assigning a Framed-IP-Address > -=09Is there a way to see the real load that is on the card (can't > be that much, PowerPC 200mhz 64mb Ram... but well you never know)=A0? not that I know of, but rest assured its not sweating much. > -=09Is the 'add defaultroute gateway xxx.yyy.zzz.nnn metric m' > really necessary for every install of a HiPerARC I don't recall explicitly telling the ARC this. You can simply give it the gateway in the QuickConfig. All a "defaultroute" is, is a gateway.....and your going to want to set a gateway on *any* tcp/ip device, including the arc. As far as the difference in: set gateway 208.206.76.1 and add defaultroute gateway 208.206.76.1 metric 1 I am sure most of us prefer the latter..........like I said, the Quickconfig should take care of the defaultroute. >=20 > Thanks for any pointers >=20 > Robert >=20 > PS=A0: Is the PPP CCP option bugged, or does it work sometimes=A0? I had = to > shut it off on all my hubs, as the ISDN connections went to hell... >=20 On the ARC? Can you explain? You only want to run CCP in some instances, something like: CCP will be attempted for call type(s): DIGITAL UNCOMPRESSED_ANALOG Brian > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >=20 /-------------------------- signal@shreve.net -----------------------------= \ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 = | | Network Administrator | Perl, Linux | Web hosting, online stores, = | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs = | | 89 CRX DX w/MPFI, lots of |-=3D*:Quake:*=3D-| http://www.shreve.net/ = | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 = | \-------------------------- 318-222-2638 x109 -----------------------------= /
Subject: Re: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: Brian <signal@shreve.net>
Date: 1998-02-18 09:14:59
On Wed, 18 Feb 1998, Pete Ashdown wrote: > It appears that either making changes to the modem settings or writing to > NVRAM causes the HiPer DSP to have a weird interaction with the ARC on ER > 4.0.80 and 4.0.72. The symptom is that the modem_group message is not > displayed and the connection's login information is not echoed back as it > is typed. > > The workaround is to do a software modem reset after you make changes to > either the modems or NVRAM. > > I am guessing the cause from some testing I did this morning. I think I > have finally settled on my settings for my DSP modems, so I don't think > I'll have to make further changes. If the bug reoccurs without me changing > anything, then I'll report back. I had some strange problems on 4.0.80 myself, specifically: 1. I had some interfaces on the hdm's go "down" for no reason, rebooting the ARC was the only way to get them to go back up. I swapped hdm's, soft/hardware reset hdm's to no avail. 2. I have to statically declare my last two hdm's as being hdm_24 with 23 ports, otherwise the ARC thinks one has 24 ports (and its a PRI!) and the other has 0 ports! 3. Still get Exception 300 reboots, and I forwarded my last crashdump to the AE guys and hopefully have a solution soon. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) MPIP on 3.7.24 and ISDN-Modem
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-18 10:01:27
Thus spake Brian >On Tue, 17 Feb 1998, Jeff Mcadams wrote: >> Thus spake Pete Ashdown >> >Practically nobody can use MPIP since we upgraded to 3.7.24 and put ISDN >> >processing on the quads. I'm going to reboot everything tomorrow to see if >> >that helps, but I'm wondering if I'm missing something in regards to MPIP >> >and ISDN-Modem processing. >> Man, 3.7.24 is the first time MPIP has worked correctly for us...we're >> running ISDN on the Munich card though....not sure how much that would >> affect it. >fwiw, mpip has almost always worked for us, but we only scale 3 chassis >with mpip, so maybe it has to do with the amount of mpip your doing. >We are running 3.7.24 with mpip working well. Yeah, that's quite possible, we're running 9 chassis in our largest hunt group, with the 7th being the control server. Of course, after I said that MPIP working yesterday, I went home last night and it was acting a bit flakey. :/ We just upgraded in this hunt group to 3.7.24 on Mon. so its still pretty new around here. I'll keep an eye on it and let you know how it holds up over time. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) converting 45amp to PSU to DC.
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-18 10:07:14
> Does anyone know the process of converting a 45amp chassis to > DC power? Just replace the power supplies. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: (usr-tc) new hiper dsp/hiper arc bundles
From: matthew <matthew@the-spa.com>
Date: 1998-02-18 10:38:53
well we have some of the new bundle on order with two hiper dsp cards, one hiper arc card etc. we haven't used the hiper arc at all and we are pretty happy with the hiper dsp cards in general so i was thinking of just taking the double up cards out of the chassis' and doubling up our other 48 port quad chassis' and worrying about using the hiper arc cards later on when i really need them and when folks say they are as reliable as the other stuff (netserver etc) so, our previous double up card kits came with the memory upgrades, the new bundles don't as they don't expect you to put it into an older chassis. what exactly are the memory chips i upgraded in the previous double up'ings and can i just buy those as standard memory/flash? does this make sense putting the double up cards in existing chassis' rather than setting up the hiper arc if i don't need it? matthew
Subject: Re: (usr-tc) new hiper dsp/hiper arc bundles
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-18 10:55:51
matthew said once upon a time: > what exactly are the memory chips i upgraded in the previous double >up'ings and can i just buy those as standard memory/flash? I might have some spares for sale sometime in the future. I think they are fairly standard, but I can't remember the spec off the top of my head. > does this make sense putting the double up cards in existing chassis' >rather than setting up the hiper arc if i don't need it? You will get *much* better performance from the ARC. The ER 4.0.72 supposedly resolves all the issues I had with the ARC. This remains to be seen as I have only been running on that version for five hours and haven't done extensive testing.
Subject: Re: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-18 10:58:28
Brian said once upon a time: >1. I had some interfaces on the hdm's go "down" for no reason, rebooting >the ARC was the only way to get them to go back up. I swapped hdm's, >soft/hardware reset hdm's to no avail. We're they "remote out of service"? I noticed this with one HDM this last weekend. Yes, I had to reboot it as well. Any attempt to put the channels in service was met with failure. The HDM code desperately needs an update. 1.0.7 is the first public release and we haven't seen anything since that came out at the end of last October! >2. I have to statically declare my last two hdm's as being hdm_24 with 23 >ports, otherwise the ARC thinks one has 24 ports (and its a PRI!) and the >other has 0 ports! Make sure they're configured properly and reboot your ARC. I had this problem as well with some new card insertions, until I did the configuration and rebooted the ARC. >3. Still get Exception 300 reboots, and I forwarded my last crashdump to >the AE guys and hopefully have a solution soon. I haven't seen this on 4.0.72 yet, but I'm watching for it.
Subject: Re: (usr-tc) Enhanced routing
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-18 11:10:03
Peter D. Mayer said once upon a time: > >I read a message here a few weeks ago about consolidating RIP routes on a >chassis into 1 large route. Just add an IP pool, add a netmask entry, turn >enh_routing on and voila! It works great, BUT: Anyone with an ip outside >that pool no longer has their ip address broadcasted via RIP. This is a >problem. I really like the enhanced routing, but does anyone know how to >make it keep broadcasting the ip's that are not in the pool? > >example: >ip pool entry: 206.175.52.1 size 61 >netmask entry: 206.175.52.0 255.255.255.192 >other ip: 205.156.197.150 255.255.255.255 (this route doesn't get >broadcasted) >Some of these ip's still get routed properly, but the route never shows up >on our Cisco. This last statement sounds a bit contradictory. Are the IPs being set up properly on the Netserver? How are you advertising routes to your Cisco?
Subject: Re: (usr-tc) Enhanced routing
From: Peter D. Mayer <dmayer@netwalk.com>
Date: 1998-02-18 11:15:18
I am using ripv2, and the ethernet address for the NETServer is outside the network of the dialup pool, but some of the static IP's I'm assigning ARE in the same network as the NETServer. Could this be causing problems? Do I maybe need to put in another netmask entry? I also have the "assigned address" and "assigned pool size" set to the same addresses as the IP pool table entry. Any ideas? Thanks, Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- >[Quoth Peter D. Mayer] >] I read a message here a few weeks ago about consolidating RIP routes on a >] chassis into 1 large route. Just add an IP pool, add a netmask entry, turn >] enh_routing on and voila! It works great, BUT: Anyone with an ip outside >] that pool no longer has their ip address broadcasted via RIP. This is a >] problem. I really like the enhanced routing, but does anyone know how to >] make it keep broadcasting the ip's that are not in the pool? >] >] example: >] ip pool entry: 206.175.52.1 size 61 >] netmask entry: 206.175.52.0 255.255.255.192 >] other ip: 205.156.197.150 255.255.255.255 (this route doesn't get >] broadcasted) >] Some of these ip's still get routed properly, but the route never shows up >] on our Cisco. > >I am doing this, and I do see routes for people who are outside of the >dialup range. Is the ethernet of your TC in a seperate network from >the IP pool? Did you turn on RIPv2 on on the TC? Try turning on RIP >debugging on the Cisco, and see if it's just rejecting them for some >reason. > >-- >Michael Mittelstadt meek@execpc.com >VP - Internet Techologies ExecPC Internet >http://www.execpc.com/~meek 1-800-ExecPC-1
Subject: Re: (usr-tc) Enhanced routing
From: Brian <signal@shreve.net>
Date: 1998-02-18 11:20:08
On Wed, 18 Feb 1998, Peter D. Mayer wrote: > I am using ripv2, and the ethernet address for the NETServer is outside the > network of the dialup pool, but some of the static IP's I'm assigning ARE in > the same network as the NETServer. Could this be causing problems? Do I yes, you really *don't* want to do that. Get some more address space, and declare a static IP pool from it. Putting static dialup users on the same network as your LAN/net0's, is going to eventually bite you. We had major problems with stale arp caches when we did this. > maybe need to put in another netmask entry? I also have the "assigned > address" and "assigned pool size" set to the same addresses as the IP pool > table entry. > > Any ideas? > > Thanks, > Peter D. Mayer > NetWalk Tech Support > dmayer@netwalk.com > > -----Original Message----- > From: Michael Mittelstadt <meek@execpc.com> > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > Date: Wednesday, February 18, 1998 12:03 AM > Subject: Re: (usr-tc) Enhanced routing > > > >[Quoth Peter D. Mayer] > >] I read a message here a few weeks ago about consolidating RIP routes on a > >] chassis into 1 large route. Just add an IP pool, add a netmask entry, > turn > >] enh_routing on and voila! It works great, BUT: Anyone with an ip > outside > >] that pool no longer has their ip address broadcasted via RIP. This is a > >] problem. I really like the enhanced routing, but does anyone know how to > >] make it keep broadcasting the ip's that are not in the pool? > >] > >] example: > >] ip pool entry: 206.175.52.1 size 61 > >] netmask entry: 206.175.52.0 255.255.255.192 > >] other ip: 205.156.197.150 255.255.255.255 (this route doesn't get > >] broadcasted) > >] Some of these ip's still get routed properly, but the route never shows > up > >] on our Cisco. > > > >I am doing this, and I do see routes for people who are outside of the > >dialup range. Is the ethernet of your TC in a seperate network from > >the IP pool? Did you turn on RIPv2 on on the TC? Try turning on RIP > >debugging on the Cisco, and see if it's just rejecting them for some > >reason. > > > >-- > >Michael Mittelstadt meek@execpc.com > >VP - Internet Techologies ExecPC Internet > >http://www.execpc.com/~meek 1-800-ExecPC-1 > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Found my 3.7.24 MPIP problem
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-18 11:43:06
Apparently upgrading to 3.7.24 erases the MPIP server configuration. Thus until I put this configuration back in, MPIP was pooched.
Subject: Re: (usr-tc) converting 45amp to PSU to DC.
From: buster_joseph/mw/us/3com@usr.com
Date: 1998-02-18 13:46:24
--IMA.Boundary.148138788 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part >> Does anyone know the process of converting a 45amp chassis to > DC power? >Just replace the power supplies. Actually, no. You have to get a new chassis. The 45amp cahssis is either Only AC or Only DC. The newer chassis with the fantray can converted.
Subject: (no subject)
From: Tod's USR-list <usrtc@ns0.ipeg.com>
Date: 1998-02-18 13:53:56
is it possible to get an isdn connection with t1 vs pri? Administrator ipeg.com
Subject: Re: (usr-tc) HELP! USR Total Control Rackmount32 Question
From: Butch Kemper <kemper@bihs.net>
Date: 1998-02-18 13:54:07
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 At 23:09 -0600 on 2/17/98, Tatai SV Krishnan wrote: > On Tue, 17 Feb 1998 BECCOM@aol.com wrote: > > > I have a total control rackmount32 and it has not been used for a while by > > another dept. We have acquired the chassis and would like to find out the > > maximum speed of these modems. The chassis holds 16 dual analog cards with > > separate RS232 nics in the back of the chassis. The NICS at the back also > > have analog ports. The chassis comes complete with RCU, dual Power >supply and > > a CIU card. To test it I connected it to an NT 4.0 server that was running > > RAS. After trying several modems, the one that worked the best was a >Courier > > v.32 w/ASL but I could only connect at 12,000 everytime not higher no >lower. > > What I am thinking is that it has some sort of initialization string >that is > > configured in the CIU and it is set for no higher than 12,000. I was >reading > > the manual and it says that I can control the Chassis by connecting the >DB9 to > > RJ45 serial connector that came included with the CIU....but I was not >aware > > of that and have to go look for it. When I entered the ATI7 command it >showed > > that it had a 16 MHz chip in it. and a few of the dates were from 93 >and 94. > > I am hoping that it is at least a 28,800 modem so we can use this box. Can > > you please let me know if there is a command that I can use to check >the speed > > or do I have to use the rj45 serial cable that was supplied to change > > initializations strings? I have worked on this for a couple of days >now and I > > am running out of options. Any support that you can supply me with >would be > > greatly appreciated. If you need to contact me please email >BECCOM@aol.com > > > > This rack is a very old rack, and I do not think that we supported 28K - > You may have only v.32 and bis modems in the chassi. > I have a similar unit in service providing 33.6 service. The key would be to pull out a modem card to see if it has a daughter board on it. When I upgraded my unit from 14.4 to 28.8, one or two daughter boards were added to each modem card. A software upgrade later added 33.6 support. Butch Butch Kemper | Free sound advice available Kemper & Associates Consulting Group | "95% sound and 5% advice" 409-361-2324 | Refunds cheerfully provided -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 5.5.3 for non-commercial use <http://www.pgp.com> iQA/AwUBNOs+iShIHyLj1QScEQIPdACgqFL8dNqg4aoTvW9HxhjCDye4Hp0Ani4L bafTgjPcZQ1x9YMLrH8eoFJr =rLL6 -----END PGP SIGNATURE-----
Subject: Re: (usr-tc) new hiper dsp/hiper arc bundles
From: matthew <matthew@the-spa.com>
Date: 1998-02-18 13:56:36
i guess i could try running one arc/dsp setup on our test pri and see how it goes. would you have any configuration tips for the arc card? matthew At 10:55 AM 2/18/98 -0700, you wrote: >matthew said once upon a time: > >> what exactly are the memory chips i upgraded in the previous double >>up'ings and can i just buy those as standard memory/flash? > >I might have some spares for sale sometime in the future. I think they are >fairly standard, but I can't remember the spec off the top of my head. > >> does this make sense putting the double up cards in existing chassis' >>rather than setting up the hiper arc if i don't need it? > >You will get *much* better performance from the ARC. The ER 4.0.72 >supposedly resolves all the issues I had with the ARC. This remains to be >seen as I have only been running on that version for five hours and haven't >done extensive testing. > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) HDM woes (again)
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-18 14:13:10
On Wed, 18 Feb 1998, Allen Marsalis wrote: > Anync PPP/Sync PPP Conversion disable <--- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This was it! After I changed this to 'disable', I was able to dial-in sync PPP w/ no problem. I am curiour however, will this affect my customers who have async ISDN adapters? (I don't have one w/ which to test) Many thanks.. :) Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-18 14:14:34
On Wed, 18 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > >1. I had some interfaces on the hdm's go "down" for no reason, rebooting > >the ARC was the only way to get them to go back up. I swapped hdm's, > >soft/hardware reset hdm's to no avail. > > We're they "remote out of service"? I noticed this with one HDM this last > weekend. Yes, I had to reboot it as well. Any attempt to put the channels > in service was met with failure. This happens to us occasionally as well. The HDMs need a good reboot in order to properly see the PRI again. Very odd. Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) [Q] Secondary IP address on net0 [Total Control
From: jason_kelton@3com.com
Date: 1998-02-18 14:26:54
I think you might be talking about virtual IP addressing - addressing multiple IP's to the one physical interface... I don't think so... The edgeserver can do it, but I'm not sure about HiPer. Regards, Jason. tracy@dwx.com on 18/02/98 13:05:09 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) Chassis]? Greetings. Excuse what is no doubt a simple question. Does anyone know an/the way to assign a secondary ip address to the net0 port of a NetServer Card? On Cisco boxes, this is simple with a command such as: ip address 192.68.123.242 255.255.255.0 secondary Does there exist such a beast for the NetServer Card? Thanks. Tracy Hinshaw, DWX Internet - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) converting 45amp to PSU to DC.
From: Henry Moats <nc0419@corp.netcom.com>
Date: 1998-02-18 14:55:06
There is an etch on the rear module for the 45 pwr supply unit that says: "for dc psu's remove etch as indicated by silkscreen" hmmmmm??? ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________| On Wed, 18 Feb 1998 Buster_Joseph/MW/US/3Com@usr.com wrote: > > > > > > > > > >> Does anyone know the process of converting a 45amp chassis to > > DC power? > > >Just replace the power supplies. > > Actually, no. You have to get a new chassis. The 45amp cahssis > is either Only AC or Only DC. > The newer chassis with the fantray can converted. > > > > ----------------------------------------------------------------- > --------- > | Curtis V. Shambeau | curt@execpc.com | > http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. > | > ----------------------------------------------------------------- > --------- > > > - > To unsubscribe to usr-tc, send an email to > "majordomo@xmission.com" with "unsubscribe usr-tc" in the body > of the message. > For information on digests or retrieving files and old messages > send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) new hiper dsp/hiper arc bundles
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-18 15:07:07
matthew said once upon a time: > > i guess i could try running one arc/dsp setup on our test pri and see how >it goes. > >would you have any configuration tips for the arc card? Go through the normal boot setup, then report back with what doesn't work. I can tell you the proper commands for your situation.
Subject: Re: (usr-tc) Need Help getting Framed user from Merit Radius
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-18 15:40:45
On Wed, 18 Mar 1998, Jeff Mcadams wrote: > Thus spake Tatai SV Krishnan > >On Wed, 18 Mar 1998, Josh Richards <Josh Richards wrote: > >> Nope - That is correct. The MTU specified in RADIUS is useless for PAP > >> based PPP logins since PPP has already started before the RADIUS server > >> even comes into the picture. > > >The above statement is correct for Livingston and NETServer - Not for > >HiPer ARC. > > How does the ARC handle this? Restart LCP negotiations? Means a > slightly longer delay in the PPP startup procedure doesn't it? If the > ARC's take half as long to auth and start PPP as the netserver do, I'm > not sure I like adding any more time to it. Not that its a *really* big > deal, but I do have customers that notice the length of time it takes to > start passing packets. :/ No HiPer ARC does not restart LCP negotiations. What happens is that you can have a user as login and network on the Hiper ARC - depending upon how he logs in the Mtu is either utilized or not. So you dialin and open a terminal and login and then join the network as a network user then the MTU is taken - if you are doing PPP as soon as you connect then the MTU from radius is not used. krish > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) Newbie questions about HiPerARC
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-18 15:47:31
Couple Newbie questions=A0: - Is there a way to replace MPIP, or to forbid MLPPP ISDN connections=A0? - How do I add a user with a fixed IP address that is not in the standard dial-in pool=A0? - Is there a way to see the real load that is on the card (can't be that much, PowerPC 200mhz 64mb Ram... but well you never know)=A0? - Is the 'add defaultroute gateway xxx.yyy.zzz.nnn metric m' really necessary for every install of a HiPerARC Thanks for any pointers Robert PS=A0: Is the PPP CCP option bugged, or does it work sometimes=A0? I = had to shut it off on all my hubs, as the ISDN connections went to hell...
Subject: Re: (usr-tc) Found my 3.7.24 MPIP problem
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-18 18:21:57
On Wed, 18 Feb 1998, Pete Ashdown wrote: > Apparently upgrading to 3.7.24 erases the MPIP server configuration. Thus > until I put this configuration back in, MPIP was pooched. > I noticed our client list in the MPIP server was gone after the updates as well. Where is that bug report form? hehe :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) Found my 3.7.24 MPIP problem
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-18 18:21:57
On Wed, 18 Feb 1998, Pete Ashdown wrote: > Apparently upgrading to 3.7.24 erases the MPIP server configuration. Thus > until I put this configuration back in, MPIP was pooched. > I noticed our client list in the MPIP server was gone after the updates as well. Where is that bug report form? hehe :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: RE: (usr-tc) Newbie questions about HiPerARC
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-18 18:37:58
Thanks for the info ! -----Original Message----- From: Brian [SMTP:signal@shreve.net] Sent: mercredi, 18. f=E9vrier 1998 16:12 To: usr-tc@lists.xmission.com Subject: Re: (usr-tc) Newbie questions about HiPerARC On Wed, 18 Feb 1998, Robert von Bismarck wrote: > - How do I add a user with a fixed IP address that is not in the > standard dial-in pool=A0? You do this in RADIUS also by assigning a Framed-IP-Address Yeah, okay, but I want to do this in the HiPerARC, so the guy connects only to this one POP, and not to all the others using the same RADIUS servers. >=20 > PS=A0: Is the PPP CCP option bugged, or does it work sometimes=A0? I had to > shut it off on all my hubs, as the ISDN connections went to hell... >=20 On the ARC? Can you explain? You only want to run CCP in some instances, something like: CCP will be attempted for call type(s): DIGITAL UNCOMPRESSED_ANALOG Brian I connected using a ZyXel Omni TA128 (very common ISDN TA around here), and had lots of CCP Resets (found out by tracing the call with the ARC monitor ppp commands) and no data (or very few) coming through at all. = I turned off CCP, and performance was up big time. I had the same problem with another TA. Never happened with analog connections. One USR tech told me to disable it, and filed the possible bug to 3com. Robert
Subject: RE: (usr-tc) Newbie questions about HiPerARC
From: Brian <signal@shreve.net>
Date: 1998-02-18 20:17:14
On Wed, 18 Feb 1998, Robert von Bismarck wrote: > Thanks for the info ! >=20 >=20 >=20 > =09-----Original Message----- > =09From:=09Brian [SMTP:signal@shreve.net] > =09Sent:=09mercredi, 18. f=E9vrier 1998 16:12 > =09To:=09usr-tc@lists.xmission.com > =09Subject:=09Re: (usr-tc) Newbie questions about HiPerARC >=20 > =09On Wed, 18 Feb 1998, Robert von Bismarck wrote: >=20 >=20 > =09> -=09How do I add a user with a fixed IP address that is not > in the > =09> standard dial-in pool=A0? >=20 > =09You do this in RADIUS also by assigning a Framed-IP-Address >=20 > Yeah, okay, but I want to do this in the HiPerARC, so the guy connects > only to this one POP, and not to all the others using the same RADIUS > servers. You can limit in RADIUS which NAS he can dial into....... >=20 > =09>=20 > =09> PS=A0: Is the PPP CCP option bugged, or does it work sometimes=A0? > I had to > =09> shut it off on all my hubs, as the ISDN connections went to > hell... > =09>=20 >=20 > =09On the ARC? Can you explain? You only want to run CCP in some > instances, > =09something like: >=20 > =09CCP will be attempted for call type(s): DIGITAL > =09 UNCOMPRESSED_ANALOG >=20 > =09Brian >=20 > I connected using a ZyXel Omni TA128 (very common ISDN TA around here), > and had lots of CCP Resets (found out by tracing the call with the ARC > monitor ppp commands) and no data (or very few) coming through at all. I > turned off CCP, and performance was up big time. I had the same problem > with another TA. Never happened with analog connections. One USR tech > told me to disable it, and filed the possible bug to 3com. >=20 can anyone else comment on this? > Robert >=20 >=20 > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >=20 /-------------------------- signal@shreve.net -----------------------------= \ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 = | | Network Administrator | Perl, Linux | Web hosting, online stores, = | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs = | | 89 CRX DX w/MPFI, lots of |-=3D*:Quake:*=3D-| http://www.shreve.net/ = | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 = | \-------------------------- 318-222-2638 x109 -----------------------------= /
Subject: Re: (usr-tc) new hiper dsp/hiper arc bundles
From: Brian <signal@shreve.net>
Date: 1998-02-18 20:18:48
On Wed, 18 Feb 1998, Pete Ashdown wrote: > matthew said once upon a time: > > > what exactly are the memory chips i upgraded in the previous double > >up'ings and can i just buy those as standard memory/flash? > > I might have some spares for sale sometime in the future. I think they are > fairly standard, but I can't remember the spec off the top of my head. > > > does this make sense putting the double up cards in existing chassis' > >rather than setting up the hiper arc if i don't need it? > > You will get *much* better performance from the ARC. The ER 4.0.72 > supposedly resolves all the issues I had with the ARC. This remains to be > seen as I have only been running on that version for five hours and haven't > done extensive testing. cool. Beings that I have 3 tickets open with 3com, on 3 different problems, and have 3 different applications engineering people working on these 3 problems, I too hope to see that code soon :) O, actually only 2 of the tickets are ARC related. Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: Brian <signal@shreve.net>
Date: 1998-02-18 20:21:10
On Wed, 18 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > >1. I had some interfaces on the hdm's go "down" for no reason, rebooting > >the ARC was the only way to get them to go back up. I swapped hdm's, > >soft/hardware reset hdm's to no avail. > > We're they "remote out of service"? I noticed this with one HDM this last > weekend. Yes, I had to reboot it as well. Any attempt to put the channels > in service was met with failure. no, I did a "list interfaces" from the ARC, and saw something like: slot:12/mod:6 down up > > The HDM code desperately needs an update. 1.0.7 is the first public > release and we haven't seen anything since that came out at the end of last > October! nod, perhaps they don't want to make a release until they have v90 ready? > > >2. I have to statically declare my last two hdm's as being hdm_24 with 23 > >ports, otherwise the ARC thinks one has 24 ports (and its a PRI!) and the > >other has 0 ports! > > Make sure they're configured properly and reboot your ARC. I had this > problem as well with some new card insertions, until I did the > configuration and rebooted the ARC. there is nothing to "configure" for the hdm on the arc really other than if your going to set them statically, and thats all i did. I know the hdm's are set right if thats what you mean. > > >3. Still get Exception 300 reboots, and I forwarded my last crashdump to > >the AE guys and hopefully have a solution soon. > > I haven't seen this on 4.0.72 yet, but I'm watching for it. > But you did see it in 4.0.80 right? 4.0.80 was suppose to fix that, but it didn't. Brian > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) HDM woes (again)
From: Brian <signal@shreve.net>
Date: 1998-02-18 20:27:03
On Wed, 18 Feb 1998, System Administrator wrote: > On Wed, 18 Feb 1998, Allen Marsalis wrote: > > > Anync PPP/Sync PPP Conversion disable <--- > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > This was it! > > After I changed this to 'disable', I was able to dial-in sync PPP w/ no > problem. I am curiour however, will this affect my customers who have > async ISDN adapters? (I don't have one w/ which to test) no..... In a PC, alot of times ISDN looks like this: AsyncPPP SyncPPP PC-------------->ISDN TA-------------------------->ISP Your pc is having a Asynchronous conversation with the TA via serial, then the TA does a Async/Sync conversion and spits it out Synchronous to talk to your USR hub. On the USR hub side, its digital from the modem to the PRI, so there is no async/sync conversion. I run Linux on my machine that dials into the net, and Linux's PPP stack is asynchronous, like most. I think in the experimental kernels they actually have synchronous ppp support, so i think I can theoretically do: SyncPPP SyncPPP Linux------------------>3com ImpactIQ------------------->ISP and eliminate the overhead of the conversion...........I know this is off topic, but can anyone comment on this? Would be nice to be sync all the way, thats one reason I like isdn routers that handle the sync ppp themselves and connect to your lan via ethernet. > > > Many thanks.. :) > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: Brian <signal@shreve.net>
Date: 1998-02-18 20:28:03
On Wed, 18 Feb 1998, System Administrator wrote: > On Wed, 18 Feb 1998, Pete Ashdown wrote: > > > Brian said once upon a time: > > > > >1. I had some interfaces on the hdm's go "down" for no reason, rebooting > > >the ARC was the only way to get them to go back up. I swapped hdm's, > > >soft/hardware reset hdm's to no avail. > > > > We're they "remote out of service"? I noticed this with one HDM this last > > weekend. Yes, I had to reboot it as well. Any attempt to put the channels > > in service was met with failure. > > This happens to us occasionally as well. The HDMs need a good reboot in > order to properly see the PRI again. Very odd. I have seen this as well. Whats screwed up, is that you can't just do a "sinserv" or something and get them back. I have seen this happen a long time ago on PRI cards as well. Brian > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Re: your mail
From: Brian <signal@shreve.net>
Date: 1998-02-18 20:29:00
On Wed, 18 Feb 1998, Tod's USR-list wrote: > > is it possible to get an isdn connection with t1 vs pri? I would imagine using DOV this would work. Brian > > Administrator > ipeg.com > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Need Help getting Framed user from Merit Radius
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-19 08:02:25
On Wed, 18 Mar 1998, Josh Richards <Josh Richards wrote: > On 18 Feb 1998, Tatai SV Krishnan wrote: > > > On Wed, 18 Mar 1998, Jeff Mcadams wrote: > > > > > Thus spake Tatai SV Krishnan > > > >On Wed, 18 Mar 1998, Josh Richards <Josh Richards wrote: > > > >> Nope - That is correct. The MTU specified in RADIUS is useless for PAP > > > >> based PPP logins since PPP has already started before the RADIUS server > > > >> even comes into the picture. > > > > > > >The above statement is correct for Livingston and NETServer - Not for > > > >HiPer ARC. > > > > > > How does the ARC handle this? Restart LCP negotiations? Means a > > > slightly longer delay in the PPP startup procedure doesn't it? If the > > > ARC's take half as long to auth and start PPP as the netserver do, I'm > > > not sure I like adding any more time to it. Not that its a *really* big > > > deal, but I do have customers that notice the length of time it takes to > > > start passing packets. :/ > > > > No HiPer ARC does not restart LCP negotiations. What happens is that you > > can have a user as login and network on the Hiper ARC - depending upon > > how he logs in the Mtu is either utilized or not. So you dialin and open > > a terminal and login and then join the network as a network user then the > > MTU is taken - if you are doing PPP as soon as you connect then the MTU > > from radius is not used. > > How is that different from how the PortMaster and 3Com NETServer handle > it? That sounds like exactly what I and several others already described. > The whole point is that you have an user who is a login/network user. In a Port master the user is either login or network. Depending upon what you choose on loggin to the HiPer ARC the setting change. krish > --jr > > ---- > Josh Richards - <jrichard@livingston.com> - [Beta Engineer] > LUCENT Technologies - Remote Access Business Unit > (formerly Livingston Enterprises, Inc.) > http://www.livingston.com/ > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) NVRAM/Hiper DSP bug with ARC
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-19 12:07:01
Brian said once upon a time: >> > >1. I had some interfaces on the hdm's go "down" for no reason, rebooting >> > >the ARC was the only way to get them to go back up. I swapped hdm's, >> > >soft/hardware reset hdm's to no avail. >> > >> > We're they "remote out of service"? I noticed this with one HDM this last >> > weekend. Yes, I had to reboot it as well. Any attempt to put the channels >> > in service was met with failure. >> >> This happens to us occasionally as well. The HDMs need a good reboot in >> order to properly see the PRI again. Very odd. > >I have seen this as well. Whats screwed up, is that you can't just do a >"sinserv" or something and get them back. I have seen this happen a long >time ago on PRI cards as well. Apparently the lack of "modem_group message" and "modem_group prompt" is a bonifide bug and has nothing to do with changing settings on the HDM modems. I'm in the process of writing a perl script to deal with this problem and the problem above.
Subject: (usr-tc) hiperhappy script
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-19 14:51:49
#!/usr/local/bin/perl # hiperhappy - a simple perl script for dealing with hiper bugs # # Rather than waiting around another month for someone to resolve my ticket # 43257, I wrote this script to reset the unused modems periodically. # Right now, I have it running every two hours, since it appears the bug is # prone to reappear within that time. # # Once I catch the PRIs going remote-out-of-service (requiring a card reboot) # I'll build in a fix for that into hiperhappy. Lets hope USR/3com fixes it # first. # # This script has no warrantees implied or denied. Use at your own risk. # pashdown@xmission.com # Be certain to change c1-23 to c1-24 if you use DSS or round-robin modems # TCM home directory $TCMHOME = "/usr/local/lib/tcm"; $ENV{'TCMHOME'} = $TCMHOME; # SNMP passwords $readsnmp = "PUBLIC"; $writesnmp = "PRIVATE"; # tcmperf and tcmcmd abbreviations $tcmperf = "$TCMHOME/bin/tcmperf -s 1 -c $readsnmp"; $tcmcmd = "$TCMHOME/bin/tcmcmd -c $readsnmp -C $writesnmp"; $| = 1; # unbuffered output # The actual call. The format is "reset_modems(NMC card,slot)". This can # be done individually or via a foreach loop as below. The routine only # resets modems that are not in use foreach $i ( 1..8 ) { &reset_modems("pop1-snmp",$i); } foreach $i ( 1..10 ) { &reset_modems("pop2-snmp",$i); } # Get out exit; # reset_modems(NMC card, slot) # resets the modems which are not in use in order to deal with the ARC's # modem_group message and modem_group prompt no-display bug sub reset_modems { ($rack, $slot) = @_; # Take the card out of service system("$tcmcmd -E \"local out of service\" -G commands $rack:s${slot}c25"); # Find unused modems print ( "$tcmperf -G \"Call Statistics\" \"Operational Status of a modem\" $rack:s${slot}c1-23\n"); open ( TCMPERF, "$tcmperf -G \"Call Statistics\" \"Operational Status of a modem\" $rack:s${slot}c1-23 2>&1 |"); while (<TCMPERF>) { if (/Operational Status of a modem (.*)/) { # Put the modem status into an array @cardstatus = split(/\s+/, $1); } } $modem = 1; $reset = ""; # Build the reset string based on unused modems while (@cardstatus) { $modemstatus = shift(@cardstatus); print ("$modem\t$modemstatus\n"); if ($modemstatus eq "idle") { if ($laststatus = "none") { $reset = $reset . "s${slot}c$modem,"; } } $modem++; } if ($reset) { system ("$tcmcmd -G \"software commands\" -E \"software reset\" $rack:$reset\n"); # Wait for the modems to reset sleep (30); } # Put the card back in service system("$tcmcmd -E \"in service\" -G commands $rack:s${slot}c25"); }
Subject: Re: (usr-tc) tcs ramblings
From: Mark van Wouw <vanwouw@gol.com>
Date: 1998-02-19 15:40:14
At 08:29 AM 2/18/98 -0600, Charles Hill wrote: >On Wed, 18 Feb 1998, Allen Marsalis wrote: > >> Well except for the lack of MPIP on the arc and OSPF anywhere, >> I'm a pretty happy camper at the moment. Never mind that I > >Ditto. > >> HDM's overall are a near perfect... My biggest complaint is that unlike the Dual PRI card, I cannot set the PRI on the HDM to do round-robin. With round-robin a wedged port/modem is a minor annoyance. First-available means many more customers are going to hit that same wedged modem. I've already had this happen. Furthermore there is NO way for me to disable or otherwise get incoming calls to skip a modem. This is a limitation of our PRI type (INS1500) I've been told. Mark. --- Global OnLine Japan - The Provider Mark van Wouw Network Operations vanwouw@gol.com 03-5341-8000 ZZ - I'm not sleepy, I just forget to escape sometimes... - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. From: Brian <signal@shreve.net> Subject: Re: (usr-tc) tcs ramblings Date: 19 Feb 1998 08:29:39 -0600 (CST) On Thu, 19 Feb 1998, Mark van Wouw wrote: > At 08:29 AM 2/18/98 -0600, Charles Hill wrote: > >On Wed, 18 Feb 1998, Allen Marsalis wrote: > > > >> Well except for the lack of MPIP on the arc and OSPF anywhere, > >> I'm a pretty happy camper at the moment. Never mind that I > > > >Ditto. > > > >> HDM's overall are a near perfect... > > My biggest complaint is that unlike the Dual PRI card, I cannot set > the PRI on the HDM to do round-robin. With round-robin a wedged > port/modem is a minor annoyance. First-available means many more > customers are going to hit that same wedged modem. I've already had > this happen. Furthermore there is NO way for me to disable or > otherwise get incoming calls to skip a modem. This is a limitation > of our PRI type (INS1500) I've been told. so "set mdmrmeth rndrobin" won't work for some PRI types? You can set up the hdm, i beleive, to busy out a modem that is screwed up, this way you don't have users getting caught up on the modem. However it makes it harder to find out when you have a problem. We use fixed assignment, which is great for troubleshooting, and we don't busy out when modem is unavailable. There are alot of pros and cons for the different routing methods, one bad thing about fixed assignment, is the user load is not spread evenly over your modems and some modems see *alot* more use than others. Brian > > Mark. > > --- > Global OnLine Japan - The Provider > Mark van Wouw Network Operations > vanwouw@gol.com 03-5341-8000 > ZZ - I'm not sleepy, I just forget to escape sometimes... > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) anyway to know the IP where client dial up
From: TieUs System Administrator <admin@tieus.com>
Date: 1998-02-19 17:29:25
This is a multi-part message in MIME format. ------=_NextPart_000_000A_01BD3D5B.EC872500 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi Is anyway I can know the IP address where client dial up from the = machine, I like to post port information from machine to the Web content! Thank you!
Subject: Re: (usr-tc) anyway to know the IP where client dial up
From: Brian <signal@shreve.net>
Date: 1998-02-19 20:39:21
On Thu, 19 Feb 1998, TieUs System Administrator wrote: > Hi > Is anyway I can know the IP address where client dial up from the machine, > I like to post port information from machine to the Web content! > I had some scripts that did this and now they are broke with HiPeR stuff. SNMP querying the modems is fine, but you have to know what port the user is on, and since ports are not mapped to IP numbers, what i did was: 1. user clicks cgi 2. we get the users ip 3. we know what hub they are on because of the ippool they were in 4. we pmwho and grep for there ip, and cut the port# and username 5. snmp query and return there info but now I cant do this, mainly because 1. pmwho doesnt work with hiperarc, and I don't have time to make it work (there is a lazyness factor there too!) 2. even if it did work, I don't know of *1* screen that shows ip number and port info and username........I don't think the same nmc queries work either. So I got the latest ucd-snmp utils, and I got the snmp.pm perl module, both latest version, thinking I was going to work on this today, but we all know things can't be *that* easy......in testing my SNMP.pm: earth:/usr/src/SNMP-1.7b2$ make test PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/lib/perl5/i586-linux/5.003 -I/usr/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t t/mib...............ok t/session...........Use of uninitialized value at blib/lib/SNMP.pm line 243. Use of uninitialized value at t/session.t line 96. Use of uninitialized value at t/session.t line 97. Use of uninitialized value at t/session.t line 97. Use of uninitialized value at t/session.t line 103. Use of uninitialized value at t/session.t line 103. FAILED tests 5-7, 9, 13 Failed 5/14 tests, 64.29% okay t/translate.........ok Failed 1 test script, 66.67% okay. 5/24 subtests failed, 79.17% okay. make: *** [test_dynamic] Error 29 So, I will have to track all this down later next week. There is also an excellent perl module called "pm???" that is like a "expect" module, lets you telnet over to something and send and receive responses, this is good for this type of thing. I hope to write something anyways, I would love to have what I had before: 1. User clicks "connect stats" on the web 2. a table comes up showing them basically most of the usefull information from a Performance Monitor 3. it would rock if it had a nice interface and worked with HiPer technology as well. Charles Hill and a few others sent me ideas/examples of CGI's which is where I got alot of the "know how", on what to do in order to get this working properly. I do know that having all calls (including ISDN) answered on the Quads (as opposed to the Munich) makes life alot easier since you can use the same snmp queries/MIB objects. I think if someone would come out with a professional "performance monitor" cgi, then all of us would use it, so our users could check there retrains, connect speed, x2 capibility etc. I am going to "Try" and get something thats portable to other isp's in the works, my last shot at this like I said, worked with quads only and only if ISDN was on the quads, and it wasnt documented very well and would have been a pain in the ass for another isp to try and customize for there setup. I *hope* someone else has already done this, would save us all alot of time, and even if its not perfect at least provide a start. Something in Java would really be cool too. Brian > Thank you! > > --------------------------- > Dan Lan > Systen Administrator of Tieus Internet Inc. > admin@tieus.com > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Netmask problem on NMC
From: John Powell <john@jetcity.com>
Date: 1998-02-19 21:32:09
I'm reconfiguring our ISP to use 4 contiguous class C networks. I entered a netmask of 255.255.252.0 into one of my NMCs. When I tried to save it to nvram, I got a message saying that the netmask was not proper for a class C address, and it wouldn't save the netmask. What can I do? Thanks! John John Powell, President john@jetcity.com Jet City Online http://www.jetcity.com Business Office: 206-281-1774 Customer Service: 425-820-7006, 1-888-747-6464
Subject: Re: (usr-tc) anyway to know the IP where client dial up
From: Michael Mittelstadt <meek@execpc.com>
Date: 1998-02-19 21:52:36
[Quoth Brian] ] On Thu, 19 Feb 1998, TieUs System Administrator wrote: ] ] > Hi ] > Is anyway I can know the IP address where client dial up from the machine, ] > I like to post port information from machine to the Web content! ] > ] ] I had some scripts that did this and now they are broke with HiPeR stuff. ] ] SNMP querying the modems is fine, but you have to know what port the user ] is on, and since ports are not mapped to IP numbers, what i did was: ] ] 1. user clicks cgi ] 2. we get the users ip ] 3. we know what hub they are on because of the ippool they were in ] 4. we pmwho and grep for there ip, and cut the port# and username ] 5. snmp query and return there info ] ] but now I cant do this, mainly because ] ] 1. pmwho doesnt work with hiperarc, and I don't have time to make it ] work (there is a lazyness factor there too!) ] ] 2. even if it did work, I don't know of *1* screen that shows ip number ] and port info and username........I don't think the same nmc ] queries work either. I'm not familiar with pmwho, but I've got some homegrown tools that take the NAS-Port and Username from RADIUS accounting and store it in a database, so I can see who is on what port. Well, with the ARC, the port numbers don't correspond to physical ports anymore. This makes taking that port number, and construction a port for SNMP querying pretty tough. I wanted to write a CGI to let users see their current rx/tx speeds, but now I can't :-( A little bird told me might change down the road. For at least the IP's, I have a program constantly read the RADIUS accounting detail files, keep an in-memory table of who is on what port, and our web server queries that when a user wants to check their time left, etc. ] So I got the latest ucd-snmp utils, and I got the snmp.pm perl module, ] both latest version, thinking I was going to work on this today, but we ] all know things can't be *that* easy......in testing my SNMP.pm: ] ] earth:/usr/src/SNMP-1.7b2$ make test ] PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib ] -I/usr/lib/perl5/i586-linux/5.003 -I/usr/lib/perl5 -e 'use Test::Harness [...] ] Failed 1 test script, 66.67% okay. 5/24 subtests failed, 79.17% okay. ] make: *** [test_dynamic] Error 29 SNMP-1.7 works great for me on Linux (RH5) with perl 5.004, even though some of the test scripts failed. I've never had a problem with it. As long as we're on the subject. Has anyone figured out how to convince the NMC to reboot a NetServer or ARC via SNMP? I couldn't figure it out from the MIBs, and I haven't gotten around to putting a packet sniffer next to our TCM box to reverse-engineer it yet. This would be really handy, because I can't run TCM on my fleet of Linux boxen, or on my firewalled Suns, and resetting wayward NetServers is about all I use TCM for most days. ] I hope to write something anyways, I would love to have what I had before: ] ] 1. User clicks "connect stats" on the web ] 2. a table comes up showing them basically most of the usefull ] information from a Performance Monitor ] 3. it would rock if it had a nice interface and worked with HiPer ] technology as well. I've done most of this with SNMP. USR's big SNMP reference that they include with each TC is a godsend. There isn't much difference between querying a HDM or a Quadmodem, and it's straightforward to write code to just ask the TC what cards it has. Leave me private mail, and I'll give you most of my program. -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1
Subject: Re: (usr-tc) anyway to know the IP where client dial up
From: David Bolen <db3l@ans.net>
Date: 1998-02-19 23:12:15
Michael Mittelstadt <meek@execpc.com> writes: > As long as we're on the subject. Has anyone figured out how to > convince the NMC to reboot a NetServer or ARC via SNMP? You can reset any slot through the chassis MIB (chs.mib) command table (uchasCmd). To reset a slot, just set the uchasCmdFunction.### object to the value 4 (hardwareReset). ### is the slot number, 1-17 for the NAC, or 21-37 for the NIC, same as that used to index the other chassis slot tables. (This is about the only place where actual slot numbers are used in the instance rather than the otherwise ubiquitous entity (slot*1000) values). Note that this is a forced reset, it's not "nice" like issuing a reboot command on the cards themselves, but it definitely does the job. The chassis command table functions as all the command tables do in the USR MIBs, so if you're being complete, after issuing the command you should query the uchasCmdResult.### object to ensure that it ends up with a successful completion (failures will have further information in uchasCmdCode.###). If you want to wait until the reset is complete, you can poll the entity operational status (uchasEntityOperStatus.####) until it enters the "operational" (4) state. During the reset, you'll see it go into "other" (1), and sometimes "failed" (5) for a short period of time before becoming operational. Note that for the entity table, the instance is the entity (slot * 1000) not just the slot number. Also, if it's a modem you're resetting, there is no card level management entity, so you need to watch one of the modems (e.g., (slot * 1000)+1). Just for the HiPer ARC, you should also be able to set the usrCfgRebootFlag object to 2 (reboot) to force the equivalent of the more graceful reboot command. You can do this either via direct SNMP to the HiPer ARC or indirectly via the NMC (so often you can use it even if you've lost IP reachability to the HiPer ARC). -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Netmask problem on NMC
From: Christopher Brown <ccbrown@feist.com>
Date: 1998-02-19 23:54:01
Not a whole lot you can do about that. The NMC won't recognize anything but a single class C. Christopher Brown ccbrown@paracom.com WAN Administrator ParaCom Technologies, Inc. > I'm reconfiguring our ISP to use 4 contiguous class C networks. I entered >a netmask of 255.255.252.0 into one of my NMCs. When I tried to save it to >nvram, I got a message saying that the netmask was not proper for a class C >address, and it wouldn't save the netmask. What can I do? > >Thanks! > >John > >John Powell, President john@jetcity.com >Jet City Online http://www.jetcity.com >Business Office: 206-281-1774 >Customer Service: 425-820-7006, 1-888-747-6464
Subject: Re: (usr-tc) Netmask problem on NMC
From: Charles Hill <chill@ionet.net>
Date: 1998-02-20 00:06:19
You can address all your NMC's in the 4th class C and use 255.255.255.0 as the netmask so the broadcast address will be correct, OR address your NMCs outside your supernet, OR wait for 3Com to fix it. I've waited for over a year so far, but I gave up a long time ago and haven't tried it with version 5.2. -CH On Thu, 19 Feb 1998, John Powell wrote: > I'm reconfiguring our ISP to use 4 contiguous class C networks. > I entered a netmask of 255.255.252.0 into one of my NMCs. When I tried > to save it to nvram, I got a message saying that the netmask was not > proper for a class C address, and it wouldn't save the netmask. What can > I do? > > Thanks! > > John > > John Powell, President john@jetcity.com > Jet City Online http://www.jetcity.com > Business Office: 206-281-1774 > Customer Service: 425-820-7006, 1-888-747-6464 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 ....
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-20 04:25:30
> I set my nmc wap ip address to 0.0.0.0 now it just keeps rebooting. What do > I do? > Reflash code to the NMC using the PCSDL. That should clear it. krish > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Slow Ping times
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-20 04:27:03
Set the modems to default using tcm - apply the hardware templte and save it nvram. krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec On Fri, 20 Feb 1998, John A Timon wrote: > Hi, before i get into this question, please realize that I have > search through the archives for a solution to this problem. > > We are haveing a unique problem with our TC box, in that when users > are connecting to one of our quad modem cards, they are getting 1200 > bps connections. when they should be getting x2 speeds :( > > this is shown to be a correct connect speed when pinging the tc box > we get 500ms + ping times. > > any ideas/suggestions/requests for more detail? > > tia > John Timon > _____________________________________________________ > "When you look at Prince Charles, don't you think > that someone in the Royal family knew someone in the > Royal family?" > --Robin Williams > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) tcs ramblings
From: Mark van Wouw <vanwouw@gol.com>
Date: 1998-02-20 10:16:37
At 08:29 AM 2/19/98 -0600, Brian wrote: >On Thu, 19 Feb 1998, Mark van Wouw wrote: >> My biggest complaint is that unlike the Dual PRI card, I cannot set >> the PRI on the HDM to do round-robin. With round-robin a wedged >> port/modem is a minor annoyance. First-available means many more >> customers are going to hit that same wedged modem. I've already had >> this happen. Furthermore there is NO way for me to disable or >> otherwise get incoming calls to skip a modem. This is a limitation >> of our PRI type (INS1500) I've been told. > >so "set mdmrmeth rndrobin" won't work for some PRI types? I stand corrected. I was trying to find where to set this for the span and found that it is a "card" value (setting through the TCM). >You can set up the hdm, i beleive, to busy out a modem that is screwed up, >this way you don't have users getting caught up on the modem. However it >makes it harder to find out when you have a problem. Right. Now that I have round robin set, busying out the modem will work. >We use fixed assignment, which is great for troubleshooting, and we don't >busy out when modem is unavailable. If you have several PRIs on the hunt group yes, if there are only a few, the percentage of busy signals a customer will get if there is a bad modem is too high IMO.
Subject: (usr-tc) _FS:Cybex commander w/extender & Cables
From: beccom@aol.com
Date: 1998-02-20 12:04:38
I have the following to sell: Cybex Commander with 8 ports Extender with 4 ports Cables come included (Cables alone are like $60.00 a piece) Let me know if anyone is interested and make offer Thanks BECCOM@aol.com
Subject: (usr-tc) NMC with wan ip set to 0.0.0.0 ....
From: Matthew Opoka <phantom@magnolia.net>
Date: 1998-02-20 13:47:04
I set my nmc wap ip address to 0.0.0.0 now it just keeps rebooting. What do I do?
Subject: (usr-tc) can't get 33.6....
From: Craig Holland <cholland@yahoo-inc.com>
Date: 1998-02-20 14:59:01
This is a multi-part message in MIME format. ------=_NextPart_000_0014_01BD3E10.14570CD0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Anyone know why I wouldn't be able to get 33.6? I can get people in at = 28.8, 56k, and isdn...but no 33.6... thanks, craig ------=_NextPart_000_0014_01BD3E10.14570CD0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN"> <HTML> <HEAD> <META content=3Dtext/html;charset=3Diso-8859-1 = http-equiv=3DContent-Type> <META content=3D'"MSHTML 4.72.2106.6"' name=3DGENERATOR> </HEAD> <BODY bgColor=3D#ffffff> <DIV><FONT color=3D#000000 size=3D2>Anyone know why I wouldn't be able = to get=20 33.6?&nbsp;&nbsp; I can get people in at 28.8, 56k, and isdn...but no=20 33.6...</FONT></DIV> <DIV><FONT color=3D#000000 size=3D2></FONT>&nbsp;</DIV> <DIV><FONT color=3D#000000 size=3D2>thanks,</FONT></DIV> <DIV><FONT color=3D#000000 size=3D2>craig</FONT></DIV></BODY></HTML> ------=_NextPart_000_0014_01BD3E10.14570CD0--
Subject: Re: (usr-tc) Slow Ping times
From: Brian <signal@shreve.net>
Date: 1998-02-20 15:33:37
On Fri, 20 Feb 1998, John A Timon wrote: > Hi, before i get into this question, please realize that I have > search through the archives for a solution to this problem. > > We are haveing a unique problem with our TC box, in that when users > are connecting to one of our quad modem cards, they are getting 1200 > bps connections. when they should be getting x2 speeds :( > > this is shown to be a correct connect speed when pinging the tc box > we get 500ms + ping times. > > any ideas/suggestions/requests for more detail? I woudn't use "ping" to test a modem connection I would use something like Performance Monitor. reset all quads to factory defaults save reset all quads to hardware profile save software reset Then you will be sure that you are good to go. Brian > > tia > John Timon > _____________________________________________________ > "When you look at Prince Charles, don't you think > that someone in the Royal family knew someone in the > Royal family?" > --Robin Williams > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Slow Ping times
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-20 15:38:06
On Fri, 20 Feb 1998, John A Timon wrote: > We are haveing a unique problem with our TC box, in that when users > are connecting to one of our quad modem cards, they are getting 1200 > bps connections. when they should be getting x2 speeds :( Saw this once on a new TC chassis. I went into TCM, do a restore to default, a save to NVRAM, and a software reset, which cured it. Brian
Subject: (usr-tc) Slow Ping times
From: John A Timon <timon@execulink.com>
Date: 1998-02-20 16:16:43
Hi, before i get into this question, please realize that I have search through the archives for a solution to this problem. We are haveing a unique problem with our TC box, in that when users are connecting to one of our quad modem cards, they are getting 1200 bps connections. when they should be getting x2 speeds :( this is shown to be a correct connect speed when pinging the tc box we get 500ms + ping times. any ideas/suggestions/requests for more detail? tia John Timon _____________________________________________________ "When you look at Prince Charles, don't you think that someone in the Royal family knew someone in the Royal family?" --Robin Williams
Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 ....
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-20 16:39:03
Tatai SV Krishnan was heard to say: >> I set my nmc wap ip address to 0.0.0.0 now it just keeps rebooting. What do >> I do? >> > >Reflash code to the NMC using the PCSDL. That should clear it. Or flip the "clear nvram" dip switch to put it back to defaults (or something similar) [I'd have to go pull one to find the right switch.] --Ricky
Subject: (usr-tc) Slow x2 response during peak times
From: Greg Coffey <greg@coffey.com>
Date: 1998-02-20 18:11:21
I got the following message from one of our subscribers today: >When I hook up at x2 late at night (after mid night) >browsing and downloading is super fast. >During primetime it is real slow unless I hook up at 33.6. This is a newer chassis, only has 24 modems running right now, we are going to 48 as soon as USWest puts in the other ds1. Any suggestions? Thanks, Greg Coffey, CoffeyNet Voice 307-234-5443 307-234-5446 Fax ~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~ 142 S. Center St. US Robotics x2 56k $20 in Casper Casper, WY 82601 Local Internet for Casper, Rawlins, Douglas, www.coffey.com Wheatland, Pinedale, Lander & Lusk, WY ~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~
Subject: Re: (usr-tc) ISDN calls and HDM's
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-20 19:50:02
On Fri, 20 Feb 1998, Robert von Bismarck wrote: > We have a customer whose Hayes Ultra 64k ISDN TA died when we made the > switch from quad modems to HDM's. We used the defaults on the modems. > Are there any differences that I should change to make this customer > happy again=A0? I had to lend him my ZyXel to keep him happy.=20 > I also have an other customer with a USR Sportster 128 who has trouble > connecting. No more info available, no negotiation, nothing. I don't > have another TA to lend him while I test the USR. >=20 > Thanks for any pointers, Not sure if this will help or not, but had a customer today who couldn't connect to HDMs w/ NT and a Motorola Bitsurfer PRO. We *finally* got it resolved, after talking to many people at Motorola. Two important things: #1. Apparently, the Bitsurfer does v120 by default. We had to disable this. #2. (still not sure on the whys of this one) In 64K sync mode, NT would refused to LCP negotiate properly (you could see lots of LCP negotiation, but final result was NT requesting disconnection). Once we put the Bitsurfer in 56K async mode, he was able to easily connect to HDMs and get both channels online. I'm still really curious why this was happening, because *clear text* (using terminal software) in 64K sync looked perfect, yet PPP wouldn't start. On the plus side, this answers my question about HDMs async->sync PPP conversion. Apparently, the HDM can autodetect when it's necessary and handle async ISDN calls (even though you do get a log message about sync_ppp failing). Regards, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: (usr-tc) ISDN calls and HDM's
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-20 20:28:10
We have a customer whose Hayes Ultra 64k ISDN TA died when we made the switch from quad modems to HDM's. We used the defaults on the modems. Are there any differences that I should change to make this customer happy again=A0? I had to lend him my ZyXel to keep him happy.=20 I also have an other customer with a USR Sportster 128 who has trouble connecting. No more info available, no negotiation, nothing. I don't have another TA to lend him while I test the USR. Thanks for any pointers, -- Robert von Bismarck Petrel Communications S.A 15, Route des Jeunes Network system engineer . Case Postale 404 Microsoft Certified Professional 1211 Gen=E8ve 24
Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 ....
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-20 20:39:58
On Sat, 21 Feb 1998, Matthew Opoka wrote: > I tried the clear CMOS jumper and also tried the dip switch 5 on. > I tried puting in a new flash chip and flashed it but it still didn't work. > It will flash USR for 1 sec then reboot and keeps doing this. > I don't know what else to try. Any other ideas. > It should take 0.0.0.0 for the wan port ip address if it's going to do this. > I think this is a major bug. Yes it is a bug, we will do address it. For now the CMOS jumper will not do any good for you. What you need is to put dip switch 5 and 6 up and flash code to it via pcsdl. Putting dip 6 on itself should clear the eeprom where the IP address is stored. krish > > > -----Original Message----- > From: Ricky Beam <jfbeam@Interpath.net> > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > Date: Friday, February 20, 1998 3:43 PM > Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 .... > > > >Tatai SV Krishnan was heard to say: > >>> I set my nmc wap ip address to 0.0.0.0 now it just keeps rebooting. > What do > >>> I do? > >>> > >> > >>Reflash code to the NMC using the PCSDL. That should clear it. > > > >Or flip the "clear nvram" dip switch to put it back to defaults (or > something > >similar) [I'd have to go pull one to find the right switch.] > > > >--Ricky > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: RE: (usr-tc) Newbie questions about HiPerARC
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-20 21:13:07
More info about the possible CCP bug : I was using my laptop, which is W95 OSR2 without dun 1.2 Connection was single-channel PPP at 64k n o MLPPP was attempted. I was running code 4.0.19, I will upgrade to 4.0.72 next week. I will check the firmware on the Zyxel and check whether this also happens with another TA and on the same software / Hardware config on the hub. I'll keep you posted on the results, Robert > > PS=A0: Is the PPP CCP option bugged, or does it work sometimes=A0? > I had to > > shut it off on all my hubs, as the ISDN connections went to > hell... > >=20 >=20 > On the ARC? Can you explain? You only want to run CCP in some > instances, > something like: >=20 > CCP will be attempted for call type(s): DIGITAL > UNCOMPRESSED_ANALOG >=20 > Brian >=20 > I connected using a ZyXel Omni TA128 (very common ISDN TA around here), > and had lots of CCP Resets (found out by tracing the call with the ARC > monitor ppp commands) and no data (or very few) coming through at all. I > turned off CCP, and performance was up big time. I had the same problem > with another TA. Never happened with analog connections. One USR tech > told me to disable it, and filed the possible bug to 3com. >=20 can anyone else comment on this? > Robert
Subject: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-21 00:38:04
---------- Forwarded message ---------- At work, we've discovered a *SERIOUS* bug in the "radius" dialup authentication software. Affected Platforms: WindowsNT (RadiusNT) Linux Solaris (x86) BSDi NetBSD OpenBSD FreeBSD Problem: If a user appends a certain amount of spaces after their username, Radius will crash, keeping users from logging in. We have been unable to determine the number of spaces, but it is above 5, and below the 'magic 128' as we call it. I'd estimate it at around 32 spaces. Effects: 100% of the time, Radius will crash. All platforms are affected. Multiple servers do not negate these effects, as most terminal servers, when the primary radius authentication server is not there, will switch over to the next one, which will get the same username, and crash, locking all customers out. This appears to affect ALL platforms, be it WindowsNT or a form of unix. It appears to be a bug in radius itself. A coworker has contacted the radius mailing lists. As soon as a fix is known, I will post it here. --Phillip R. Jaenke (prj@raex.com | prj@nls.net) Primary Developer, The Improvement Linux Project Core Team Member, The Cyberian RC5 Effort - http://www.cyberian.org/ AKA Kaeyerai (Rediscovery) of MasterTechnoMonster Ketyra Designs, Inc. - Imagine Transmeta sans Linus. That's us. :)
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-21 03:09:29
Laszlo Vecsey was heard to say: >---------- Forwarded message ---------- >Date: Fri, 20 Feb 1998 21:02:53 -0500 >From: "Phillip R. Jaenke" <prj@NLS.NET> >To: BUGTRAQ@NETSPACE.ORG >Subject: Serious bug in "radius" dialup authentication software > >At work, we've discovered a *SERIOUS* bug in the "radius" dialup >authentication software. ... >Problem: >If a user appends a certain amount of spaces after their username, Radius >will crash, keeping users from logging in. We have been unable to >determine the number of spaces, but it is above 5, and below the 'magic >128' as we call it. I'd estimate it at around 32 spaces. It's not crashing my RADIUS server(s) -- USR Security/Accounting v4.3.11. (Granted, there are other ways to crash it, but spaces isn't one of them.) It looks like a bug with your specific radius server. --Ricky
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-21 03:09:29
Laszlo Vecsey was heard to say: >---------- Forwarded message ---------- >Date: Fri, 20 Feb 1998 21:02:53 -0500 >From: "Phillip R. Jaenke" <prj@NLS.NET> >To: BUGTRAQ@NETSPACE.ORG >Subject: Serious bug in "radius" dialup authentication software > >At work, we've discovered a *SERIOUS* bug in the "radius" dialup >authentication software. ... >Problem: >If a user appends a certain amount of spaces after their username, Radius >will crash, keeping users from logging in. We have been unable to >determine the number of spaces, but it is above 5, and below the 'magic >128' as we call it. I'd estimate it at around 32 spaces. It's not crashing my RADIUS server(s) -- USR Security/Accounting v4.3.11. (Granted, there are other ways to crash it, but spaces isn't one of them.) It looks like a bug with your specific radius server. --Ricky
Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 ....
From: Matthew Opoka <phantom@magnolia.net>
Date: 1998-02-21 03:38:30
I tried the clear CMOS jumper and also tried the dip switch 5 on. I tried puting in a new flash chip and flashed it but it still didn't work. It will flash USR for 1 sec then reboot and keeps doing this. I don't know what else to try. Any other ideas. It should take 0.0.0.0 for the wan port ip address if it's going to do this. I think this is a major bug. -----Original Message----- >Tatai SV Krishnan was heard to say: >>> I set my nmc wap ip address to 0.0.0.0 now it just keeps rebooting. What do >>> I do? >>> >> >>Reflash code to the NMC using the PCSDL. That should clear it. > >Or flip the "clear nvram" dip switch to put it back to defaults (or something >similar) [I'd have to go pull one to find the right switch.] > >--Ricky > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Brian <signal@shreve.net>
Date: 1998-02-21 09:02:53
I am not so sure I buy this. How could this go unchecked in *all* versions of radius? Not to mention, when RADIUS "crashes" or exits, its restarted by inetd anyways........ On Sat, 21 Feb 1998, Laszlo Vecsey wrote: > ---------- Forwarded message ---------- > Date: Fri, 20 Feb 1998 21:02:53 -0500 > From: "Phillip R. Jaenke" <prj@NLS.NET> > To: BUGTRAQ@NETSPACE.ORG > Subject: Serious bug in "radius" dialup authentication software > > At work, we've discovered a *SERIOUS* bug in the "radius" dialup > authentication software. > > Affected Platforms: > WindowsNT (RadiusNT) > Linux > Solaris (x86) > BSDi > NetBSD > OpenBSD > FreeBSD > > Problem: > If a user appends a certain amount of spaces after their username, Radius > will crash, keeping users from logging in. We have been unable to > determine the number of spaces, but it is above 5, and below the 'magic > 128' as we call it. I'd estimate it at around 32 spaces. > > Effects: > 100% of the time, Radius will crash. All platforms are affected. Multiple > servers do not negate these effects, as most terminal servers, when the > primary radius authentication server is not there, will switch over to the > next one, which will get the same username, and crash, locking all > customers out. This appears to affect ALL platforms, be it WindowsNT or a > form of unix. It appears to be a bug in radius itself. > > A coworker has contacted the radius mailing lists. As soon as a fix is > known, I will post it here. > > --Phillip R. Jaenke (prj@raex.com | prj@nls.net) > Primary Developer, The Improvement Linux Project > Core Team Member, The Cyberian RC5 Effort - http://www.cyberian.org/ > AKA Kaeyerai (Rediscovery) of MasterTechnoMonster > Ketyra Designs, Inc. - Imagine Transmeta sans Linus. That's us. :) > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-21 09:37:21
On Sat, 21 Feb 1998, Laszlo Vecsey wrote: Heh. Interesting to note that the original author NEVER actually specifies WHICH radius server he is talking about! Doesn't say anything about USR, Livingston, Merit, etc. Maybe it's an exploit only found in "Joe's RADIUS server"? :) Or perhaps he discovered a hear-to unread footnote in the RFC stating that "A RADIUS server implementation MUST dump core if it receives 32 spaces as a username."? Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. "It tastes like chicken, Charlie Brown" -- New Peanuts TV special. The whole Peanuts gang struggles to survive high in the Andes mountains after their privately chartered plane goes down due to pilot error. > ---------- Forwarded message ---------- > Date: Fri, 20 Feb 1998 21:02:53 -0500 > From: "Phillip R. Jaenke" <prj@NLS.NET> > To: BUGTRAQ@NETSPACE.ORG > Subject: Serious bug in "radius" dialup authentication software > > At work, we've discovered a *SERIOUS* bug in the "radius" dialup > authentication software. > > Affected Platforms: > WindowsNT (RadiusNT) > Linux > Solaris (x86) > BSDi > NetBSD > OpenBSD > FreeBSD > > Problem: > If a user appends a certain amount of spaces after their username, Radius > will crash, keeping users from logging in. We have been unable to > determine the number of spaces, but it is above 5, and below the 'magic > 128' as we call it. I'd estimate it at around 32 spaces. > > Effects: > 100% of the time, Radius will crash. All platforms are affected. Multiple > servers do not negate these effects, as most terminal servers, when the > primary radius authentication server is not there, will switch over to the > next one, which will get the same username, and crash, locking all > customers out. This appears to affect ALL platforms, be it WindowsNT or a > form of unix. It appears to be a bug in radius itself. > > A coworker has contacted the radius mailing lists. As soon as a fix is > known, I will post it here.
Subject: (usr-tc) Version 3.02 - slow speeds?
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-21 13:52:04
Ever since we upgraded to Version 3.02 for the TCH (we have the 2059 bundle, quad digital modem cards, etc.), we have had a number of customers complain about slow connection speeds through the TC. Has anyone else had this problem?
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-21 17:11:21
On Sat, 21 Feb 1998, Laszlo Vecsey wrote: > ---------- Forwarded message ---------- > Date: Fri, 20 Feb 1998 21:02:53 -0500 > From: "Phillip R. Jaenke" <prj@NLS.NET> > To: BUGTRAQ@NETSPACE.ORG > Subject: Serious bug in "radius" dialup authentication software > > At work, we've discovered a *SERIOUS* bug in the "radius" dialup > authentication software. > > Affected Platforms: > WindowsNT (RadiusNT) > Linux > Solaris (x86) > BSDi > NetBSD > OpenBSD > FreeBSD This warning is pretty worthless without the name of the Radius software being used, and the version number or numbers. I really doubt that every version of Radius on every platform will stop responding when this string of characters is sent. Brian
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-21 17:52:48
Brian was heard to say: > >I am not so sure I buy this. How could this go unchecked in *all* >versions of radius? Not to mention, when RADIUS "crashes" or exits, its >restarted by inetd anyways........ > Unless inetd thinks it is looping and then it gets shutdown for 5minutes. --Ricky
Subject: (usr-tc) Re: Serious bug in "radius" dialup authentication software (fwd)
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-21 19:06:12
More clarification, again from the original author :) ---------- Forwarded message ---------- >You're not telling us which radius server. Livingston 1.16 or 2.01? >Merit? Cistron? etc (As a matter of fact I am sure Cistron is safe). Since this is the 22nd email I've recieved on this, I decided to CC: to bugtraq so everyone will PLEASE stop asking me this. So far, tested servers are: Livingston 1.16 to 2.01 RadiusNT v2.x Merit So far, the only one NOT vulnerable is Merit. Cistron is untested, so I've got not idea whether or not it is. Best way to test is to telnet to a terminal server, and login with a valid username, with 40 or more spaces after it. As to Cistron being safe; safe is really relative here. If somebody nasty has your dialup numbers, then you might have to restart radius a lot. Otherwise, there's really no security risk that I've found. -prj -Ed Kuchar (InterNIC Handle: EK113) [ekuchar@NLS.NET] NetLink Services, Inc. 216.468.5100(Cleveland) - 330.940.2700(Akron) sales@nls.net - http://www.nls.net - http://www.getinfo.net Serving: Cleveland, Akron, Medina, & Geauga County
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Brian <signal@shreve.net>
Date: 1998-02-21 19:45:21
On Sat, 21 Feb 1998, Ricky Beam wrote: > Brian was heard to say: > > > >I am not so sure I buy this. How could this go unchecked in *all* > >versions of radius? Not to mention, when RADIUS "crashes" or exits, its > >restarted by inetd anyways........ > > > > Unless inetd thinks it is looping and then it gets shutdown for 5minutes. 1. If inetd *really* does loop, then you do have some serious problems that need fixing, but I have never seen RADIUS loop. 2. Normally inetd will shutdown when it receives more than 40 requests in a 60 second period. You can up this by setting the "max" directive in /etc/inetd.conf. We have done this on our mail server for the pop3 service, and its probably defintily a good idea for most ISP's to up this on the RADIUS side. Brian > > --Ricky > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Serious bug in "radius" dialup authentication software (fwd)
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-21 22:50:53
Thus spake Brian >I am not so sure I buy this. How could this go unchecked in *all* >versions of radius? Not to mention, when RADIUS "crashes" or exits, its >restarted by inetd anyways........ This, of course, assumes that you're running radiusd out of inetd, which we've found no need to do. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Weird Critical Event
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-22 09:05:34
On Sat, 21 Mar 1998, Scott Kreuser wrote: > > Hi all. > > New to this list just setup our first TC rack.. Now, it seems > everyone is logging in ok but I'm getting the following critical > event whenever someone logs out. > > At 18:35:29, Facility "Call Initiation Process", Level "CRITICAL":: CIP: > User ca > ryn is not configured as dial out user. > Set the type to dial_out in the user record > How is the user caryn configured? krish > Any clues?? > > I don't have any users trying to dial out. > > Scott > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Chan T1 monitor "RT>"
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-22 09:21:15
On Sun, 22 Feb 1998, Brian wrote: > On Sun, 22 Feb 1998, Mark R. Lindsey wrote: > > > Occasionally when I'm connected to one of my chanellized T1 cards > > through the serial console line, I'll hit the wrong key, and I'm > > thrown into some sort of ROM monitor or some such with an "RH>" or > > "RT>" prompt. There's some online help, but I can't seem to get out of > > the bloomin' thing. Does anybody know how to exit it and get back to the > > normal operating mode? > > > > Thanks. > > On newer code releases it is | to get into RT> and out of RT>, on older > releases I believe this may have been ~ but I am unsure. It is ~. krish > > Control-D (debug menu) > and | > > are two of the "hidden" undocumented areas in the PRI card, mostly used > by 3com engineers. > > > > > --- > > Mark R. Lindsey, mark@datasys.net > > Internet Engineer, DSS Online > > Voice: +1 912 241 0607; Fax: +1 912 241 0190 > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) Problems activating new modem card
From: Lars Freund <lafreund@cip.e-technik.uni-erlangen.de>
Date: 1998-02-22 13:17:07
Hi USR-people, I'm having a problem with installing a new QuadModemCard in a Network Express Hub. I plugged the card in the chassis, configured the modems, activated the Slot for QuadModem in the Dual Pri Card, stored everything into the NVRAM but the calls only go to the old modem cards, not to the new one. The olds have slot 2 to 4, the new one is in slot 5. After rebooting the netserver I was surprised: the QuadModem-Slot5 was deleted in the Dual Pri Card and the logs told me not to find anything in Slot 5: Feb 21 01:24:36 polyp Chassis Awareness notice: slot# 0, all channels, NAC type = 27, new status 1 Feb 21 01:24:36 polyp Chassis Awareness notice: slot# 1, all channels, NAC type = 13, new status 1 Feb 21 01:24:36 polyp Chassis Awareness notice: slot# 2, all channels, NAC type = 13, new status 1 Feb 21 01:24:37 polyp Chassis Awareness notice: slot# 3, all channels, NAC type = 13, new status 1 Feb 21 01:24:37 polyp Chassis Awareness notice: slot# 4, all channels, NAC type = 13, new status 1 Feb 21 01:24:37 polyp Chassis Awareness notice: slot# 5, all channels, NAC type = 0, new status 0 This happened after another reboot of the NetServer Card as well, so it is impossible that I have forgotten to store the nvram. ^^^^^^^^^^^^^^^^^^ So I told the Dual Pri Card that Slot 5 (s17-s20) is a QuadModem again, but it still doesn't work. The modem reports being "present" now, but the calls only go from S5 to S16. Have you got any idea what I had forgotten? Bye, Lars Chassis Slot Device Configuration Status Device Slot# Type 1 Dual-PRI 2 Qbch-mdm 3 Qbch-mdm 4 Qbch-mdm 5 Qbch-mdm 6 NONE 7 NONE 8 NONE 9 NONE 10 NONE 11 NONE 12 NONE 13 NONE 14 NONE 15 NONE 16 ISDN-GW Chassis Slot Device Configuration Status Current Configuration Status Device Device Slot# Type Slot# Type 1 Dual-PRI 9 NONE 2 Qbch-mdm 10 NONE 3 Qbch-mdm 11 NONE 4 Qbch-mdm 12 NONE 5 Qbch-mdm 13 NONE 6 NONE 14 NONE 7 NONE 15 NONE 8 NONE 16 ISDN-GW Command> sh ses Port User Host/Inet/Dest Type Dir Status Start Idle ---- --------------- ---------------- ------- --- ------------- ------ ------ S0 - - Log/Net In USERNAME - 84days S1 - - Login In IDLE 0 0 S2 - - Login In IDLE 0 0 S3 - - Login In IDLE 0 0 S4 - - Login In IDLE 0 0 S5 fo0005 polyp06.forchhei Netwrk In ESTABLISHED 8 0 S6 fo0146 polyp04.forchhei Netwrk In ESTABLISHED 22 0 S7 - - Log/Net In IDLE 0 0 S8 fo0631 194.95.197.176 Netwrk In ESTABLISHED 2:25 0 S9 - - Log/Net In IDLE 6 0 S10 fo0207 polyp11.forchhei Netwrk In ESTABLISHED 7 0 S11 fo0378 194.95.197.165 Netwrk In ESTABLISHED 35 0 S12 fo0178 194.95.197.163 Netwrk In ESTABLISHED 51 0 S13 fo0110 194.95.197.173 Netwrk In ESTABLISHED 33 0 S14 fo5005 194.95.197.170 Netwrk In ESTABLISHED 19 0 S15 fo0551 194.95.197.172 Netwrk In ESTABLISHED 31 0 S16 fo0346 polyp29.forchhei Netwrk In ESTABLISHED 5 0 -- Press Return for More -- S17 - - Log/Net In IDLE 0 0 S18 - - Log/Net In IDLE 0 0 S19 - - Log/Net In IDLE 0 0 S20 - - Log/Net In IDLE 0 0 S21 - - Login In IDLE 0 0 S22 - - Login In IDLE 0 0 S23 - - Login In IDLE 0 0 S24 - - Login In IDLE 0 0 Command> sh s17 ----------------------- Current Status - Port S17 --------------------------- Status: IDLE Input: 0 Parity Errors: 0 Output: 0 Framing Errors: 0 Pending: 0 Overrun Errors: 0 Active Configuration Default Configuration (* = Host -------------------- --------------------- Can Override) Port Type: Login/Netwrk Login/Netwrk (Dial In) (Security) Login Service: PortMux PortMux Device Service: Telnet@7017 Telnet@7017 Modem Stat: READY ACTIVE Databits: 8 8 Stopbits: 1 1 Parity: none none Flow Control: None None* Modem Control: on on HDLC Framing: PPP in modem PPP in modem SLIP Framing: SLIP in NETServer SLIP in NETServer -- Press Return for More -- Modem Presence: Modem is present Init Script: (None) Init When?: Never Hosts: default Terminal Type: Login Prompt: $hostname login: Idle Timeout: 10 min 0 sec Dial Group: 0 TAP type: OFF OFF -- +-----------------------------------------------------------------+ | Lars Freund lars@forchheim.com | | FOnLine Buergernetzverein: http://www.forchheim.baynet.de | | http://cip2.e-technik.uni-erlangen.de:8080/hyplan/lafreund.html | +-----------------------------------------------------------------+
Subject: (usr-tc) Chan T1 monitor "RT>"
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-22 13:32:37
Occasionally when I'm connected to one of my chanellized T1 cards through the serial console line, I'll hit the wrong key, and I'm thrown into some sort of ROM monitor or some such with an "RH>" or "RT>" prompt. There's some online help, but I can't seem to get out of the bloomin' thing. Does anybody know how to exit it and get back to the normal operating mode? Thanks. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190
Subject: (usr-tc) bsdcomp
From: Brian <signal@shreve.net>
Date: 1998-02-22 14:35:05
Will the "bsdcomp" that comes with the net distributed pppd-2.x.x implementations work with Total Control in any way? Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: RE: (usr-tc) Weird Critical Event
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-22 14:41:26
On Sun, 22 Mar 1998, Scott Kreuser wrote: > Standard RADIUS authentication from a UNIX box. > check it out I remmed the three lines out below, > now I'm not getting the error. But, I still > don't see why I'm getting the error when they are > not remmed out. I was looking in some of the > docs and they said PORT-LIMIT was not supported. > Maybe thats the problem. But how do you prevent users > from logging in twice?? Port-Limit is supported. I am not aware of any doc that says port-limit is not supported. The three lines that you have commented out has nothing to do with a dial out user or the error. Are you sure that if you uncomment these three lines the error does reappear? krish > > Scott > > DEFAULT Password = "UNIX" > User-Service-Type = Framed-User, > Framed-Protocol = PPP, > Framed-Address = 255.255.255.254, > Framed-Netmask = 255.255.255.255, > Framed-Routing = None, > Framed-Compression = Van-Jacobsen-TCP-IP, > Framed-Filter-Id = "std.ppp.in", > Framed-MTU = 1500 > # Idle-Timeout = 1200, > # Session-Timeout = 28800, > # Port-Limit = 1 > > > > > Hi all. > > > > > > New to this list just setup our first TC rack.. Now, it seems > > > everyone is logging in ok but I'm getting the following critical > > > event whenever someone logs out. > > > > > > At 18:35:29, Facility "Call Initiation Process", Level "CRITICAL":: CIP: > > > User ca > > > ryn is not configured as dial out user. > > > Set the type to dial_out in the user record > > > > > How is the user caryn configured? > > > > krish > > > > > Any clues?? > > > > > > I don't have any users trying to dial out. > > > > > > Scott > > > > > > >
Subject: Re: (usr-tc) Chan T1 monitor "RT>"
From: Brian <signal@shreve.net>
Date: 1998-02-22 17:54:24
On Sun, 22 Feb 1998, Mark R. Lindsey wrote: > Occasionally when I'm connected to one of my chanellized T1 cards > through the serial console line, I'll hit the wrong key, and I'm > thrown into some sort of ROM monitor or some such with an "RH>" or > "RT>" prompt. There's some online help, but I can't seem to get out of > the bloomin' thing. Does anybody know how to exit it and get back to the > normal operating mode? > > Thanks. On newer code releases it is | to get into RT> and out of RT>, on older releases I believe this may have been ~ but I am unsure. Control-D (debug menu) and | are two of the "hidden" undocumented areas in the PRI card, mostly used by 3com engineers. > > --- > Mark R. Lindsey, mark@datasys.net > Internet Engineer, DSS Online > Voice: +1 912 241 0607; Fax: +1 912 241 0190 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Going Crazy
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-22 22:45:00
A little over a week ago I posted a message about the problem I was having getting access to the Security and Accounting RADIUS server software on the TotalService website. I also sent a lengthy description about the very strange conversation I had with one of the support personnel about the supposed need to repurchase the software each year in addition to the regular software maintenance on the chassis and other software. I want to report back that 3Com came through and aknowledged that I wasn't going crazy and that the person I spoke to misunderstood the 3Com policies. I received a very cordial letter from the person who looked into this and she promprtly got things taken care of. I suspect Krish forwarded her my message. Anyway, I know 3Com gets bad press in here when things go astray and I wanted to make sure they also got credit for setting things straight. Now if the concurrency bug is fixed in this release, you can consider me one very happy camper. Jeff Binkley ASA Network Computing
Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 ....
From: Matthew Opoka <phantom@magnolia.net>
Date: 1998-02-22 23:44:41
Tried what you said and still no go just a constant reboot after the USR show on the front of the card. I think I'm going to call and get another card from my vendor. -----Original Message----- Cc: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> >On Sat, 21 Feb 1998, Matthew Opoka wrote: > >> I tried the clear CMOS jumper and also tried the dip switch 5 on. >> I tried puting in a new flash chip and flashed it but it still didn't work. >> It will flash USR for 1 sec then reboot and keeps doing this. >> I don't know what else to try. Any other ideas. >> It should take 0.0.0.0 for the wan port ip address if it's going to do this. >> I think this is a major bug. > >Yes it is a bug, we will do address it. For now the CMOS jumper will not >do any good for you. What you need is to put dip switch 5 and 6 up and >flash code to it via pcsdl. Putting dip 6 on itself should clear the >eeprom where the IP address is stored. > >krish > >> >> >> -----Original Message----- >> From: Ricky Beam <jfbeam@Interpath.net> >> To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> >> Date: Friday, February 20, 1998 3:43 PM >> Subject: Re: (usr-tc) NMC with wan ip set to 0.0.0.0 .... >> >> >> >Tatai SV Krishnan was heard to say: >> >>> I set my nmc wap ip address to 0.0.0.0 now it just keeps rebooting. >> What do >> >>> I do? >> >>> >> >> >> >>Reflash code to the NMC using the PCSDL. That should clear it. >> > >> >Or flip the "clear nvram" dip switch to put it back to defaults (or >> something >> >similar) [I'd have to go pull one to find the right switch.] >> > >> >--Ricky >> > >> >- >> > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> > with "unsubscribe usr-tc" in the body of the message. >> > For information on digests or retrieving files and old messages send >> > "help" to the same address. Do not use quotes in your message. >> > >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the message. >> For information on digests or retrieving files and old messages send >> "help" to the same address. Do not use quotes in your message. >> > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Version 3.02 - slow speeds?
From: Joe Zulanas <joez@met-net.com>
Date: 1998-02-23 09:22:09
At 01:52 PM 2/21/98 -0600, you wrote: > >Ever since we upgraded to Version 3.02 for the TCH (we have the 2059 >bundle, quad digital modem cards, etc.), we have had a number of customers >complain about slow connection speeds through the TC. Has anyone else had >this problem? > > We have had this problem when they connect it is @ 26.4 or slightly less. We also get the problem they will connect @ 44 and only get ~ 2 to 3 K throughput even on the same internal network. Has anybody seen this problem? Joseph Zulanas CEO __________________________________________________________ The Metropolitan Network, Inc To The Internet & Beyond! We supply you with all the Internet you can eat, for just $19.95. __________________________________________________________
Subject: Re: (usr-tc) Version 3.02 - slow speeds?
From: Joe Zulanas <joez@met-net.com>
Date: 1998-02-23 09:22:09
At 01:52 PM 2/21/98 -0600, you wrote: > >Ever since we upgraded to Version 3.02 for the TCH (we have the 2059 >bundle, quad digital modem cards, etc.), we have had a number of customers >complain about slow connection speeds through the TC. Has anyone else had >this problem? > > We have had this problem when they connect it is @ 26.4 or slightly less. We also get the problem they will connect @ 44 and only get ~ 2 to 3 K throughput even on the same internal network. Has anybody seen this problem? Joseph Zulanas CEO __________________________________________________________ The Metropolitan Network, Inc To The Internet & Beyond! We supply you with all the Internet you can eat, for just $19.95. __________________________________________________________
Subject: (usr-tc) Dwindling Connections
From: Jose de Leon <jadiel@thevision.net>
Date: 1998-02-23 09:46:25
I'm having a strange problem with X2 connections. Since I'm new to this list, maybe it has already been solved. I have several customers, (not all), who connect at respectable X2 speeds and within a period of two minutes there is a continuous downgrade in speed until eventually the link is considered dead the modem hangs up. I've tried the simple things like verifying hardware flow control on client side is set and port bps is correct on client side. I'm using 5.8.6 on the quads, 3.7.24 on the netserver, 5.2.2 on the NMC. Any ideas? Thanks, Jose
Subject: (usr-tc) Couriers connecting to TC at x2
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-23 10:41:24
I have a Courier modem with x2. Recently, I started seeing 1000s of BLERs when connected to a TC rack with modem code 5.6.7. I thought the modem was bad, so I swapped it for another Courier running the 02/28/97 code. The errors went away. Then I upgraded the code to the 7/97 code in the Courier. I started to get 1000s of errors again. Downgraded to the 2/97 code, and it all works again. Any ideas if this is a Courier issue, or a TC rack issue? I haven't tried upgrading the modem code on the TC rack. Should I do that? I was just going to wait for the v.90 code for my next upgrade, but maybe I shouldn't wait. Brian
Subject: (usr-tc) Couriers connecting to TC at x2
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-23 10:41:24
I have a Courier modem with x2. Recently, I started seeing 1000s of BLERs when connected to a TC rack with modem code 5.6.7. I thought the modem was bad, so I swapped it for another Courier running the 02/28/97 code. The errors went away. Then I upgraded the code to the 7/97 code in the Courier. I started to get 1000s of errors again. Downgraded to the 2/97 code, and it all works again. Any ideas if this is a Courier issue, or a TC rack issue? I haven't tried upgrading the modem code on the TC rack. Should I do that? I was just going to wait for the v.90 code for my next upgrade, but maybe I shouldn't wait. Brian
Subject: Re: (usr-tc) Dwindling Connections
From: Brian <signal@shreve.net>
Date: 1998-02-23 12:13:33
On Mon, 23 Feb 1998, Jose de Leon wrote: > I'm having a strange problem with X2 connections. Since I'm new to this > list, maybe it has already been solved. > > I have several customers, (not all), who connect at respectable X2 speeds > and within a period of two minutes there is a continuous downgrade in speed > until eventually the link is considered dead the modem hangs up. > > I've tried the simple things like verifying hardware flow control on client > side is set and port bps is correct on client side. > > I'm using 5.8.6 on the quads, 3.7.24 on the netserver, 5.2.2 on the NMC. > > Any ideas? Sounds like typical "spiral death syndrom". Make sure the client code is the latest x2 code. It probably isn't. Brian > > Thanks, > Jose > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Off-Topic...venting
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-23 12:22:26
You know....I *used* to *like* Pachalbel's Canon. :) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) Off-Topic...venting
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-23 12:22:26
You know....I *used* to *like* Pachalbel's Canon. :) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Netserver idle-timeout
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-23 12:26:50
Richard Bosire said once upon a time: >I have this problem of netserver idle-timeout ,, The suppossedly >working feature does not work on my setup .. Are you using RADIUS to set the Idle-Timeout? This has worked for my setup for a while.
Subject: Re: (usr-tc) Netserver idle-timeout
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-23 12:26:50
Richard Bosire said once upon a time: >I have this problem of netserver idle-timeout ,, The suppossedly >working feature does not work on my setup .. Are you using RADIUS to set the Idle-Timeout? This has worked for my setup for a while.
Subject: (usr-tc) interesting lack of communication
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-23 12:47:32
From what I can tell, the tech support droid type of people that you talk to after listen to Pachalbel's Canon for the 85th time for the most part from what I can tell don't consider the (very common given the code that I've seen) possibility that there might be a bug in the code. On the last call I placed concerning the binary mode telnet, I made an almost off-hand comment that I thought it was a bug in the code, and the tech asked if I wanted to open a bug report (apparently, just opening a case just isn't enough for something like this), so I said yes, they transfer'ed me somewhere else where they open'ed a bug report (apparently seperate from the case number) to be dealt with by Engineering or R&D or some group like that. Anyway...word to the wise that apparently, just opening up a case with tech support doesn't necessarily get the word to the right people about a bug in the code. :/ -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) interesting lack of communication
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-23 12:47:32
From what I can tell, the tech support droid type of people that you talk to after listen to Pachalbel's Canon for the 85th time for the most part from what I can tell don't consider the (very common given the code that I've seen) possibility that there might be a bug in the code. On the last call I placed concerning the binary mode telnet, I made an almost off-hand comment that I thought it was a bug in the code, and the tech asked if I wanted to open a bug report (apparently, just opening a case just isn't enough for something like this), so I said yes, they transfer'ed me somewhere else where they open'ed a bug report (apparently seperate from the case number) to be dealt with by Engineering or R&D or some group like that. Anyway...word to the wise that apparently, just opening up a case with tech support doesn't necessarily get the word to the right people about a bug in the code. :/ -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Chan T1 monitor "RT>"
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-23 12:48:05
Hi .. Have you tried the obvious .. like exit,done,quit ?. cheers bosire Mark R. Lindsey wrote: > Occasionally when I'm connected to one of my chanellized T1 cards > through the serial console line, I'll hit the wrong key, and I'm > thrown into some sort of ROM monitor or some such with an "RH>" or > "RT>" prompt. There's some online help, but I can't seem to get out of > the bloomin' thing. Does anybody know how to exit it and get back to the > normal operating mode? > > Thanks. > > --- > Mark R. Lindsey, mark@datasys.net > Internet Engineer, DSS Online > Voice: +1 912 241 0607; Fax: +1 912 241 0190 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) Chan T1 monitor "RT>"
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-23 12:48:05
Hi .. Have you tried the obvious .. like exit,done,quit ?. cheers bosire Mark R. Lindsey wrote: > Occasionally when I'm connected to one of my chanellized T1 cards > through the serial console line, I'll hit the wrong key, and I'm > thrown into some sort of ROM monitor or some such with an "RH>" or > "RT>" prompt. There's some online help, but I can't seem to get out of > the bloomin' thing. Does anybody know how to exit it and get back to the > normal operating mode? > > Thanks. > > --- > Mark R. Lindsey, mark@datasys.net > Internet Engineer, DSS Online > Voice: +1 912 241 0607; Fax: +1 912 241 0190 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: (usr-tc) Netserver idle-timeout
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-23 12:51:23
Hi .. I have this problem of netserver idle-timeout ,, The suppossedly working feature does not work on my setup .. Is anyone having the same problem , is there any work-around .. my netserver details are: U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.3.28 Build date: Dec 13 1996 Build time: 13:54:59 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Enhanced Licensed for 60 ports. ] cheers and thanx -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: (usr-tc) 1200 and 2400 baud connections
From: John A Timon <timon@execulink.com>
Date: 1998-02-23 14:07:18
Hi all, I am having a problem with connecting. I have 3 chassis, 2 with 12 quads, and 1 with 12 quads and 2 hiper dsp cards. all three with i dentical configurations. The problem is on one of the chassis I have a quad that is continually connecting at 1200 baud, I have tried reseting the modem to default and writing to nvram and software rebooting the card to no avail, am I missing something? Also i have another chasis that every modem in it connects at 2400 baud everytime. it has the exact configuration as the others, and i have tried restoring the default there too, to no avail. has anyone seen this, and/or does anyone possible have a solution. tia John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: (usr-tc) 1200 and 2400 baud connections
From: John A Timon <timon@execulink.com>
Date: 1998-02-23 14:07:18
Hi all, I am having a problem with connecting. I have 3 chassis, 2 with 12 quads, and 1 with 12 quads and 2 hiper dsp cards. all three with i dentical configurations. The problem is on one of the chassis I have a quad that is continually connecting at 1200 baud, I have tried reseting the modem to default and writing to nvram and software rebooting the card to no avail, am I missing something? Also i have another chasis that every modem in it connects at 2400 baud everytime. it has the exact configuration as the others, and i have tried restoring the default there too, to no avail. has anyone seen this, and/or does anyone possible have a solution. tia John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-23 15:45:18
Laszlo Vecsey said once upon a time: > >Now that the new netserver code supposedly improves the quake lag >situation, can users return to a default init string? Or is there still >significant benefit by having the user turn off compression etc. Only bad >thing is they need to dial up again when doing anything else but Quake, >such as surfing the web, to get maximum performance. Users can use whatever string they like. I have turned off V.42 SREJ on all my equipment though. I have had very good Quake results with the new code. Both Quake2 and QuakeWorld play much better than before. I would say that getting ISDN processing off the Munich board is one of the big keys towards eliminating lag. I still get a periodic bump, or a bad connection, but things aren't hopeless like they used to be.
Subject: (usr-tc) Quakeworld and Quake2 lag
From: Laszlo Vecsey <master@internexus.net>
Date: 1998-02-23 16:47:05
Now that the new netserver code supposedly improves the quake lag situation, can users return to a default init string? Or is there still significant benefit by having the user turn off compression etc. Only bad thing is they need to dial up again when doing anything else but Quake, such as surfing the web, to get maximum performance. - lv
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian <signal@shreve.net>
Date: 1998-02-23 17:09:42
On Mon, 23 Feb 1998, Pete Ashdown wrote: > Laszlo Vecsey said once upon a time: > > > >Now that the new netserver code supposedly improves the quake lag > >situation, can users return to a default init string? Or is there still > >significant benefit by having the user turn off compression etc. Only bad > >thing is they need to dial up again when doing anything else but Quake, > >such as surfing the web, to get maximum performance. > > Users can use whatever string they like. I have turned off V.42 SREJ on > all my equipment though. I thought Sel Rej *helped* with quake.........your saying you disable selrej, to help quake? > > I have had very good Quake results with the new code. Both Quake2 and > QuakeWorld play much better than before. I would say that getting ISDN > processing off the Munich board is one of the big keys towards eliminating > lag. I still get a periodic bump, or a bad connection, but things aren't > hopeless like they used to be. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-23 17:44:50
Thus spake Laszlo Vecsey >Now that the new netserver code supposedly improves the quake lag >situation, can users return to a default init string? Or is there still >significant benefit by having the user turn off compression etc. Only bad >thing is they need to dial up again when doing anything else but Quake, >such as surfing the web, to get maximum performance. Turning off compression etc. will pretty much *always* improve latency issues...so if you have any programs that are latency sensitive, you will get better performance for them with compression turned off and such. This was not just a workaround for this USR specific situation, twas merely a way of mitigating the situation a bit by eeking every bit of latency improvement out of the connection. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Dwindling Connections
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-23 18:22:34
On Mon, 23 Feb 1998, Brian wrote: > On Mon, 23 Feb 1998, Jose de Leon wrote: > > > I'm having a strange problem with X2 connections. Since I'm new to this > > list, maybe it has already been solved. > > > > I have several customers, (not all), who connect at respectable X2 speeds > > and within a period of two minutes there is a continuous downgrade in speed > > until eventually the link is considered dead the modem hangs up. > > > > I've tried the simple things like verifying hardware flow control on client > > side is set and port bps is correct on client side. > > > > I'm using 5.8.6 on the quads, 3.7.24 on the netserver, 5.2.2 on the NMC. > > > > Any ideas? > > > Sounds like typical "spiral death syndrom". Make sure the client code is > the latest x2 code. It probably isn't. Actually, I have seen this a few times lately. I have seen it most often on the HDM's, but thought it was a client issue or just a temporary thing. I have seen it once on netserver 3.7.24 with 5.8.6 quad code, and 3 or 4 times on HDM 1.0.7 with Arc 4.0.95. The thing is, the clients have varied from Courier v.everything w/X2, to Sportster 33.6. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-23 20:58:08
On Mon, 23 Feb 1998, Laszlo Vecsey wrote: > Now that the new netserver code supposedly improves the quake lag > situation, can users return to a default init string? Or is there still I think supposedly is the correct word here. I am still seeing quake problems with the latest stuff. Only, instead of lag, my customers are having problems with getting disconnected, but only while playing quake. They don't get disconnected with MP/16s hooked to Portmasters. Brian
Subject: Re: (usr-tc) mdmCsStatus
From: Ian Roy <iroy@smokies.amnix.com>
Date: 1998-02-23 22:28:08
Brian, Could you shoot me an e-mail once you have it working ? Thanks Ian Roy > I am writing a program so I can tell how many users are logged on all my > hubs at any given time, if nothing else, I am also trying to get an > education in ucd SNMP. > > I am querying mdmCsStatus to find out if the modem is in use, let me know > if there is a better way. > > My program is not working, here is what I have: > > > use SNMP; > SNMP::setMib("/usr/local/mibs/mib.txt"); > > $host = shift; > unless ($host) { > $| = 1; print "enter SNMP host address: "; $| = 0; > chomp($host = <STDIN>); > } > > $secret = shift; > unless ($secret) { > $| = 1; print "enter SNMP community string: "; $| = 0; > chomp($secret = <STDIN>); > } > > $session = new SNMP::Session ( DestHost => $host, Community => $secret) || die; > > for ($i=0; $i<48; $i++) { > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > $session->get($mdmCsStatus); > print "> $mdmCsStatus\n"; > } > > I think my problem has to do with this line: > > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > > Does anyone know what I could be doing wrong? Thanks. > > Also if anyone has already patched pmwho to give some kind of useful > output when used with a ARC that would be great and save me some time, > otherwise thats my next project. > > Brian > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > -- ******************************************************************************** ** ** ** #include <std/disclaimer.h> ** ** ** ** Ian Roy E-Mail: iroy@amnix.com ** ** AMNIX Communications WWW: http://www.amnix.com ** ** 2993 S. Peoria St., Suite 128 Phone: (303) 337-2878 ** ** Aurora, CO 80014 Fax: (303) 743-0923 ** ** ** ********************************************************************************
Subject: (usr-tc) mdmCsStatus
From: Brian <signal@shreve.net>
Date: 1998-02-23 23:17:34
I am writing a program so I can tell how many users are logged on all my hubs at any given time, if nothing else, I am also trying to get an education in ucd SNMP. I am querying mdmCsStatus to find out if the modem is in use, let me know if there is a better way. My program is not working, here is what I have: use SNMP; SNMP::setMib("/usr/local/mibs/mib.txt"); $host = shift; unless ($host) { $| = 1; print "enter SNMP host address: "; $| = 0; chomp($host = <STDIN>); } $secret = shift; unless ($secret) { $| = 1; print "enter SNMP community string: "; $| = 0; chomp($secret = <STDIN>); } $session = new SNMP::Session ( DestHost => $host, Community => $secret) || die; for ($i=0; $i<48; $i++) { $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); $session->get($mdmCsStatus); print "> $mdmCsStatus\n"; } I think my problem has to do with this line: $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); Does anyone know what I could be doing wrong? Thanks. Also if anyone has already patched pmwho to give some kind of useful output when used with a ARC that would be great and save me some time, otherwise thats my next project. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 00:28:48
Brian <signal@shreve.net> writes: > for ($i=0; $i<48; $i++) { > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > $session->get($mdmCsStatus); > print "> $mdmCsStatus\n"; > } > > I think my problem has to do with this line: > > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > > Does anyone know what I could be doing wrong? Thanks. You're using the wrong instance on the MIB object. I'm not familiar with this particular SNMP package, but presuming thtat the Varbind is just suffixing the value of $i onto your MIB object, that's definitely the wrong value. The NMC addresses all manageable entities within the chassis through the use of an "entity" notation. The entity for a manageable object is (<slot>*1000)+unit. Unit can be 0 for card level management entities. For modems, each modem is a unit off of the slot entity (there isn't actually a manageable slot entity for a quad card). So for the quad card in slot 2, the entities are 2001,2002,2003,2004. For slot 2, it's 3001, 3002, 3003, 3004, and so on. HDMs work the same way but have units 1-24 or 1-30 depending on T1/E1. The one other quirk with HDMs is for the DS1 entities - unlike a dual card where the DS1s are unit 1 and 2, because the HDM has modems there, it numbers the DS1s just after the modems, so for an HDM in slot 14, the DS1 is either 14025 (T1 HDM) or 14031 (E1 HDM). You can either change your loop to initialize according to the appropriate instance, or - since you're already performing individual Gets with the associated inefficiency (it's faster to poll multiple modems per queries, although you have to deal with failures within a single object of the PDU), you could just use GetNext to walk any modems in the chassis. For example, if you initialized to mdmCsStatus, and then kept doing GetNext until you got an object that wasn't beneath mdmCsStatus, you'd walk all the modems in turn, skipping over any empty slots (or problem modems for that matter). -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 00:31:07
(I wrote) > So for the quad > card in slot 2, the entities are 2001,2002,2003,2004. For slot 2, > it's 3001, 3002, 3003, 3004, and so on. Argh - the second case should obviously start "For slot 3". -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: (usr-tc) TC Hub HiPer Arc IP allocation problems
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 02:52:59
We just set up a new TC hub with the HiPer Arc router card. We currently are using some RADIUS servers to do our authentication (AimTraveler RADIUS running on NT). Everything looks fine for a while, but then the HiPer Arc card starts stealing IP addresses from the rest of our network. We've verified this by matching the MAC address of the HiPer Arc with the hardware address in the NT log files. The RADIUS log files, however, show that the IP addresses given to the callers are from the IP Address pool in the HiPer Arc configuration. Does anyone understand what I am talking about (it's very late and I've been working on this for 3 straight days)? Does anyone have any clues on where to fix this bug? I really need to get the new hub up and running ASAP. -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 02:55:12
On Tue, 24 Feb 1998, Mike wrote: > At 01:34 PM 2/24/98 -0600, you wrote: > >BTW, USR told me that you need to set the IP pools to 15% more IP > >addresses than the system needs to service the modems in your hub. > >Something to do with the time needed to release the IP addresses back > >into the pool upon disconnection. > > This is only true if you are using "hint assigned" or PPTP. If you are not > using these set > your limit to the correct value. > It is only true if you are using Hint assigned - If you are using PPTP then the address is given out by NT RAS so you need to setup the pool on NT not on NETServer. If you are using Hint assigned on the Hiper ARC we do suggest to increase your pool by 2 or more addresses. krish > >Mike wrote: > >> > >> After you set up the pools and the limit did you save all & reboot? The ip > >> pool and limit only take effect after > >> a reboot.. This has caused many people some grief.. > >> > > `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' > Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics > Network Systems Engineer > PGP: http://coredump.ae.usr.com/pgp (Prefered) > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 03:53:58
On Tue, 24 Feb 1998, Andrew Aken wrote: > What is "hint assigned" > Hint assigned is a way of telling the radius server what IP to give when a user dials in. This is something that you have to do by modifying the radius server. What happens is when a user comes in, and is authenticated by Radius, the NETServer/HiPer arc can send the Radius and IP address from its pool that will be assigned to the user. This method is called hint assigned. krish > Tatai SV Krishnan wrote: > > > > It is only true if you are using Hint assigned - If you are using PPTP > > then the address is given out by NT RAS so you need to setup the pool on > > NT not on NETServer. > > > > If you are using Hint assigned on the Hiper ARC we do suggest to increase > > your pool by 2 or more addresses. > > > > krish > > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 05:02:34
FWIW, the new HiPer ARC card fixed the Quake lag problem for me. Unfortunately, because of some other bugs in the code, I can't use it... Pete Ashdown wrote: > > Laszlo Vecsey said once upon a time: > > > >Now that the new netserver code supposedly improves the quake lag > >situation, can users return to a default init string? Or is there still > >significant benefit by having the user turn off compression etc. Only bad > >thing is they need to dial up again when doing anything else but Quake, > >such as surfing the web, to get maximum performance. > > Users can use whatever string they like. I have turned off V.42 SREJ on > all my equipment though. > > I have had very good Quake results with the new code. Both Quake2 and > QuakeWorld play much better than before. I would say that getting ISDN > processing off the Munich board is one of the big keys towards eliminating > lag. I still get a periodic bump, or a bad connection, but things aren't > hopeless like they used to be. > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 06:51:44
On Tue, 24 Feb 1998, Andrew Aken wrote: > Are there any settings on the TCH which I must make to use "hint > assigned" addressing? set hint on/off Now there is nothing on the TCH which will make use of this - it is designed for Radius use. krish > > Tatai SV Krishnan wrote: > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > What is "hint assigned" > > > > > > > Hint assigned is a way of telling the radius server what IP to give when > > a user dials in. This is something that you have to do by modifying the > > radius server. What happens is when a user comes in, and is > > authenticated by Radius, the NETServer/HiPer arc can send the Radius and > > IP address from its pool that will be assigned to the user. This method > > is called hint assigned. > > > > krish > > > > > ======================================================= > > > =========== Andrew Aken - President ========= > > > ====== GlobalEyes Communications, Inc. ====== > > > =Southern Illinois' Fastest Connection to the Internet= > > > ========== http://www.GlobalEyes.net ======== > > > ======================================================= > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the message. > > > For information on digests or retrieving files and old messages send > > > "help" to the same address. Do not use quotes in your message. > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Netserver idle-timeout
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-24 07:36:16
Pete Ashdown wrote: > Richard Bosire said once upon a time: > > >I have this problem of netserver idle-timeout ,, The suppossedly > >working feature does not work on my setup .. > > Are you using RADIUS to set the Idle-Timeout? This has worked for my setup > > for a while. > Hi Pete ,No , I have been trying it from the netserver command prompt .. Pliz you dont mind shedding some ideas on how to do this from RADIUS TIA cheers bosire > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) mdmCsStatus
From: Brian <signal@shreve.net>
Date: 1998-02-24 07:53:36
> > Hi Brian .. > > I ma not sure , wether it is the same output you want, but i have to small c > program that telnets to the netserver and does a show session and output the stuff > to your screen > > The program pwho ( shareware) is a completely c prog and easy to install .. > > cheers > > bosire I have pmwho, but wanted to use snmp, and I want to eventually have something that will work with HiPer as well. > > > > > > > /-------------------------- signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > > | Network Administrator | Perl, Linux | Web hosting, online stores, | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > -- > > \\|// - ? > (o o) > +==================================oOOo=(_)=oOOo========+ > | Richard Bosire rbosire@africaonline.co.ke | > | AfricaOnline Ltd | > | union towers, 2nd floor | > | tel: 254-2-243775 | > | .oooO | > | http://www.africaonline.co.ke ( ) Oooo. | > +===================================\ (==( )==========+ > \_) ) / > (_/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Off-Topic...venting
From: Brian <signal@shreve.net>
Date: 1998-02-24 07:55:10
On Mon, 23 Feb 1998, Jeff Mcadams wrote: > You know....I *used* to *like* Pachalbel's Canon. :) Didnt they have Rondo alla Turka and a few other songs on that loop as well? I spoke to someone there about the music, they said its like so embedded into there phone system its a major bitch to change. Brian > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian <signal@shreve.net>
Date: 1998-02-24 07:56:57
On Tue, 24 Feb 1998, Andrew Aken wrote: > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > Unfortunately, because of some other bugs in the code, I can't use it... which bugs specifically? > > Pete Ashdown wrote: > > > > Laszlo Vecsey said once upon a time: > > > > > >Now that the new netserver code supposedly improves the quake lag > > >situation, can users return to a default init string? Or is there still > > >significant benefit by having the user turn off compression etc. Only bad > > >thing is they need to dial up again when doing anything else but Quake, > > >such as surfing the web, to get maximum performance. > > > > Users can use whatever string they like. I have turned off V.42 SREJ on > > all my equipment though. > > > > I have had very good Quake results with the new code. Both Quake2 and > > QuakeWorld play much better than before. I would say that getting ISDN > > processing off the Munich board is one of the big keys towards eliminating > > lag. I still get a periodic bump, or a bad connection, but things aren't > > hopeless like they used to be. > > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Why?
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 08:05:37
You are using an IP subnet mask of 255.255.255.0 for the user. Change the IP subnet mask to 255.255.255.255 for host based addressing or use a subnet mask that you want the user to have if he has a network attached. krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec On Mon, 23 Mar 1998, Scott Kreuser wrote: > > argghhh. > > I setup a user on my TC and when I dialed in to test he was connected > to another term server somewhere and I got a IP address conflict on > my TC. I disconnected him from the other term server, and now > we are still getting IP Address conflicts when he tries to log in. > I know that this IP does not exist any where on my network! > > At 23:12:01, Facility "IP", Level "CRITICAL":: User mediat is configured for > an > existing IP network address (d006b800). > > > Anyone seen this? > > Scott > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Off-Topic...venting
From: Brian <signal@shreve.net>
Date: 1998-02-24 08:27:09
> > Yeah, but you have to get past Pachalbel's Canon first...one and a half > times since the loop is screwed up and it starts over about halfway > through the piece. And with hold times getting lower as they have been > (I do have to give them credit for that) I never get past the Canon > anymore. Yes, I am definitly happy with hold times these days. > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) mdmCsStatus
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-24 08:30:26
Brian wrote: > I am writing a program so I can tell how many users are logged on all my > hubs at any given time, if nothing else, I am also trying to get an > education in ucd SNMP. > > I am querying mdmCsStatus to find out if the modem is in use, let me know > if there is a better way. > > My program is not working, here is what I have: > > use SNMP; > SNMP::setMib("/usr/local/mibs/mib.txt"); > > $host = shift; > unless ($host) { > $| = 1; print "enter SNMP host address: "; $| = 0; > chomp($host = <STDIN>); > } > > $secret = shift; > unless ($secret) { > $| = 1; print "enter SNMP community string: "; $| = 0; > chomp($secret = <STDIN>); > } > > $session = new SNMP::Session ( DestHost => $host, Community => $secret) || die; > > for ($i=0; $i<48; $i++) { > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > $session->get($mdmCsStatus); > print "> $mdmCsStatus\n"; > } > > I think my problem has to do with this line: > > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > > Does anyone know what I could be doing wrong? Thanks. > > Also if anyone has already patched pmwho to give some kind of useful > output when used with a ARC that would be great and save me some time, > otherwise thats my next project. > > Brian Hi Brian .. I ma not sure , wether it is the same output you want, but i have to small c program that telnets to the netserver and does a show session and output the stuff to your screen The program pwho ( shareware) is a completely c prog and easy to install .. cheers bosire > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) mdmCsStatus
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-24 08:30:26
Brian wrote: > I am writing a program so I can tell how many users are logged on all my > hubs at any given time, if nothing else, I am also trying to get an > education in ucd SNMP. > > I am querying mdmCsStatus to find out if the modem is in use, let me know > if there is a better way. > > My program is not working, here is what I have: > > use SNMP; > SNMP::setMib("/usr/local/mibs/mib.txt"); > > $host = shift; > unless ($host) { > $| = 1; print "enter SNMP host address: "; $| = 0; > chomp($host = <STDIN>); > } > > $secret = shift; > unless ($secret) { > $| = 1; print "enter SNMP community string: "; $| = 0; > chomp($secret = <STDIN>); > } > > $session = new SNMP::Session ( DestHost => $host, Community => $secret) || die; > > for ($i=0; $i<48; $i++) { > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > $session->get($mdmCsStatus); > print "> $mdmCsStatus\n"; > } > > I think my problem has to do with this line: > > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > > Does anyone know what I could be doing wrong? Thanks. > > Also if anyone has already patched pmwho to give some kind of useful > output when used with a ARC that would be great and save me some time, > otherwise thats my next project. > > Brian Hi Brian .. I ma not sure , wether it is the same output you want, but i have to small c program that telnets to the netserver and does a show session and output the stuff to your screen The program pwho ( shareware) is a completely c prog and easy to install .. cheers bosire > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) mdmCsStatus
From: Brian <signal@shreve.net>
Date: 1998-02-24 08:30:45
On Tue, 24 Feb 1998, David Bolen wrote: > Brian <signal@shreve.net> writes: > > > for ($i=0; $i<48; $i++) { > > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > > $session->get($mdmCsStatus); > > print "> $mdmCsStatus\n"; > > } > > > > I think my problem has to do with this line: > > > > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > > > > Does anyone know what I could be doing wrong? Thanks. > > You're using the wrong instance on the MIB object. I'm not familiar > with this particular SNMP package, but presuming thtat the Varbind is > just suffixing the value of $i onto your MIB object, that's definitely > the wrong value. > > The NMC addresses all manageable entities within the chassis through > the use of an "entity" notation. The entity for a manageable object > is (<slot>*1000)+unit. Unit can be 0 for card level management > entities. Understood. But how would I have known this? I tried to read the MIB and make sense of it, and read thru the Parameter Reference. Take someone like me, who wants to embark on writing some SNMP stuff for the TC, how do I figure out that the modem entities are (<slot>*1000)+unit? Is this documented somewhere, if so I will read it............I think thats my first problem is not having all the information I need at hand. > > -- David > > /-----------------------------------------------------------------------\ > \ David Bolen \ Internet: db3l@ans.net / > | ANS Communications, Inc. \ Phone: (914) 701-5327 | > / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ > \-----------------------------------------------------------------------/ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) mdmCsStatus
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-24 08:38:00
David, Is there a target number of modems per query you use to avoid exceeding the PDU length limit of the NMC or possibly a fragmentation problem ? I've not looked at the length of an individual modem query PDU nor the reply packet. Jeff Binkley ASA Network Computing u>Brian <signal@shreve.net> writes: u>> for ($i=0; $i<48; $i++) { u>> $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); u>> $session->get($mdmCsStatus); u>> print "> $mdmCsStatus\n"; u>> } u>> I think my problem has to do with this line: u>> $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); u>> Does anyone know what I could be doing wrong? Thanks. u>You're using the wrong instance on the MIB object. I'm not familiar u>with this particular SNMP package, but presuming thtat the Varbind is u>just suffixing the value of $i onto your MIB object, that's definitely u>the wrong value. u>The NMC addresses all manageable entities within the chassis through u>the use of an "entity" notation. The entity for a manageable object u>is (<slot>*1000)+unit. Unit can be 0 for card level management u>entities. u>For modems, each modem is a unit off of the slot entity (there isn't u>actually a manageable slot entity for a quad card). So for the quad u>card in slot 2, the entities are 2001,2002,2003,2004. For slot 2, u>it's 3001, 3002, 3003, 3004, and so on. u>HDMs work the same way but have units 1-24 or 1-30 depending on u>T1/E1. The one other quirk with HDMs is for the DS1 entities - unlike u>a dual card where the DS1s are unit 1 and 2, because the HDM has u>modems there, it numbers the DS1s just after the modems, so for an HDM u>in slot 14, the DS1 is either 14025 (T1 HDM) or 14031 (E1 HDM). u>You can either change your loop to initialize according to the u>appropriate instance, or - since you're already performing individual u>Gets with the associated inefficiency (it's faster to poll multiple u>modems per queries, although you have to deal with failures within a u>single object of the PDU), you could just use GetNext to walk any u>modems in the chassis. u>For example, if you initialized to mdmCsStatus, and then kept doing u>GetNext until you got an object that wasn't beneath mdmCsStatus, you'd u>walk all the modems in turn, skipping over any empty slots (or problem u>modems for that matter). u>-- David u>/--------------------------------------------------------------------- u>--\ \ David Bolen \ Internet: u>db3l@ans.net / | ANS Communications, Inc. \ u>Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 u>\ Fax: (914) 701-5310 \ u>\---------------------------------------------------------------------- - u>/ u>- u> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" u> with "unsubscribe usr-tc" in the body of the message. u> For information on digests or retrieving files and old messages send u> "help" to the same address. Do not use quotes in your message. u> CMPQwk 1.42 9999
Subject: Re: (usr-tc) mdmCsStatus
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 08:58:09
At 11:17 PM 2/23/98 -0600, you wrote: >I am writing a program so I can tell how many users are logged on all my >hubs at any given time, if nothing else, I am also trying to get an >education in ucd SNMP. > >I am querying mdmCsStatus to find out if the modem is in use, let me know >if there is a better way. > Brian, This will work but it requires a SNMP get for each modem in the chassis.. This is slow.. A better method is to query the Chassis LED's.. This would be two gets per chassis (one for on/off and one for color if you only want to count the calls that are connected and "green").. The OID's for this were in a message I sent a long time ago to the list.. Check the archives for them.. If you cant find them I will dig them up and repost it.. When using this on a HDSP card you cant get an exact count of calls up, but you can get a pretty good estimate by multiplying the number of lit LED's by 2.3 for PRI (10 leds * 2.3 = 23 calls) or 2.4 for T1. -m
Subject: Re: (usr-tc) 1200 and 2400 baud connections
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 09:00:39
At 02:07 PM 2/23/98 +0000, you wrote: >Hi all, I am having a problem with connecting. I have 3 chassis, 2 >with 12 quads, and 1 with 12 quads and 2 hiper dsp cards. all three >with i dentical configurations. The problem is on one of the chassis >I have a quad that is continually connecting at 1200 baud, I have >tried reseting the modem to default and writing to nvram and software >rebooting the card to no avail, am I missing something? Also i have >another chasis that every modem in it connects at 2400 baud >everytime. it has the exact configuration as the others, and i have >tried restoring the default there too, to no avail. > >has anyone seen this, and/or does anyone possible have a solution. Are you saving to NVRAM and then doing a hardware reset? The other thing to check is that your default settings are still there after the reset.. If they are not you might have "autoconfig on card init" turned on on your NMC.. Turn off autoconfig and repeat your set to default,save and reboot.. -m `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)
Subject: Re: (usr-tc) Couriers connecting to TC at x2
From: chris loelke <chris@okanagan.net>
Date: 1998-02-24 09:02:21
i have been struggling with this problem since we switched to TC. i have had USR looking into this since january. I can call the socrates server (usr bbs) without problems (0-30 BLERS. If i call our own TC, i get about 1100 BLERS per minute. Sometimes (1 out of 10), i get zero BLERS. If i get a zero BLER connection, i can stay on for hours without a problem.... The other thing i noticed was the the SNR (signal to noise ratio) is around 36 db when i get BLERS, but around 50 db when i get zero BLERS. This same modem, when used at a different exchange has no problems at all. Some of our other users mainly X2, are having the same problem, but they complain that it is slow to use the system. This issue has been escalated to level 2 at USR, i am trying to escalate this to level 999999. At 10:41 AM 2/23/98 -0600, you wrote: >I have a Courier modem with x2. Recently, I started seeing 1000s of BLERs >when connected to a TC rack with modem code 5.6.7. > >I thought the modem was bad, so I swapped it for another Courier running >the 02/28/97 code. The errors went away. Then I upgraded the code to the >7/97 code in the Courier. I started to get 1000s of errors again. >Downgraded to the 2/97 code, and it all works again. > >Any ideas if this is a Courier issue, or a TC rack issue? > >I haven't tried upgrading the modem code on the TC rack. Should I do >that? I was just going to wait for the v.90 code for my next upgrade, but >maybe I shouldn't wait. > >Brian > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Off-Topic...venting
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-24 09:05:40
Thus spake Brian >On Mon, 23 Feb 1998, Jeff Mcadams wrote: >> You know....I *used* to *like* Pachalbel's Canon. :) >Didnt they have Rondo alla Turka and a few other songs on that loop as >well? I spoke to someone there about the music, they said its like so >embedded into there phone system its a major bitch to change. Yeah, but you have to get past Pachalbel's Canon first...one and a half times since the loop is screwed up and it starts over about halfway through the piece. And with hold times getting lower as they have been (I do have to give them credit for that) I never get past the Canon anymore. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) mdmCsStatus
From: Brian <signal@shreve.net>
Date: 1998-02-24 09:19:53
I am abandoning my idea of checking mdmCsStatus because of its inefficiencies noted by David and Michael. But it was enough to get my feet wet with UCD SNMP, as opposed to CMU SNMP. CMU SNMP is no longer supported by the SNMP perl module. Thanks for the help, and thanks to michael mittlestadt for giving me good examples in ucd snmp to get my bearings, Here is a simple program using the snmp perl module that inefficiently queries all modems for there status in slots 2-13 (quads only). I am just putting it here so that others can see something simple: #!/usr/bin/perl # mdmstatus # <signal@shreve.net> # 23 February 1998 # v0.1 use SNMP; SNMP::setMib("/usr/local/mibs/mib.txt"); $status = { 1 => 'Idle', 2 => 'Off Hook', 3 => 'Dialing', 4 => 'Ringing', 5 => 'Ring Received', 6 => 'Link Negotiation', 7 => 'Online Originate', 8 => 'Online Answer', 9 => 'Local Command Mode', 10 => 'Remote Command Mode', 11 => 'Line Busied Out', 12 => 'Tone Test', 13 => 'Responder Test 105', 14 => 'Responder Test 104', 33 => 'Failed', 49 => 'Testing ROM', 50 => 'Testing RAM', 51 => 'Testing NVRAM', 52 => 'Analog Loopback', 53 => 'Local Digital Loopback', 54 => 'Remote Digital Loopback', 55 => 'Self Test', 56 => 'Phone Test', 57 => 'Non Managed Device', 58 => 'Slot Empty', 59 => 'Modem Disabled', 60 => 'Not Responding', 61 => 'Non Managable Device', 62 => 'MIU Failed', }; $host = shift; unless ($host) { $| = 1; print "enter SNMP host address: "; $| = 0; chomp($host = <STDIN>); } $secret = shift; unless ($secret) { $| = 1; print "enter SNMP community string: "; $| = 0; chomp($secret = <STDIN>); } $session = new SNMP::Session ( DestHost => $host, Community => $secret) || die; for ($o=2000; $o<14000; $o=$o+1000) { for ($i=1; $i<5; $i++) { $slot=$o/1000; $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $o+$i]); $session->get($mdmCsStatus); print "Slot $slot Modem $i $$status{$mdmCsStatus->[2]}\n"; } } I will search for the OID's mw was talking about in the archives, LED's and guestimation should be good enough. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) archives
From: Brian <signal@shreve.net>
Date: 1998-02-24 09:33:10
Where is the USR-TC archives at, to search on older messages. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Jose de Leon <jadiel@thevision.net>
Date: 1998-02-24 09:44:22
I'm not use ARC, and I get the same problem, the technician from USR suggested I enable proxy arp. I just don't think it should be necessary. The interface should not assign an IP address that is not specified for the system. -----Original Message----- On Tue, 24 Feb 1998, Andrew Aken wrote: > I posted another message earlier regarding the bug we are experiencing > here with the HiPer ARC. Basically, whenever anyone logs onto the hub > with the HiPer ARC, it starts assigning random addresses to the MAC on > the HiPer ARC. These addresses are usually from one of my servers which > then renders the server unreachable. Usually within 20 minutes of the > new hub being on-line. What version of code are you using? It must be something specific your doing, as I have never seen this posted here before, and I know there are alot of people using ARC's without this problem. We have run the release code and 3 or 4 ER codes and never saw it "proxyarp" server addresses to our MAC of our ARC. > > But, while it was on-line, our Quake players were back to the 120 pings > like we had earlier. BTW, on our Total Control Hub with the Netserver > card, we did not have any ping problems until we added the HiPer DSP's > to the chassis. It was just too much of a load for the Netserver. > Brian > Brian wrote: > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > > > Unfortunately, because of some other bugs in the code, I can't use it... > > > > which bugs specifically? > > > > > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/ - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: (usr-tc) restricting connect rates...
From: Jay Nitikman <jay@cruzio.com>
Date: 1998-02-24 09:45:38
I'm sorry if this has been covered before but I cannot find the answer in the archives. I want to restrict modem connections to X2 only. Below is a dump from Total Control Manager of the "Signal Convert Settings" for the "Modem Programmed Settings". Can you explain to me what I need to change inorder to restrict dialups to X2? Link Rate Speed Select (&N) variable Non-ARQ Transmit Buffer Size (S15.3) bytes1500 bytes1500 Buffer RX During MNP Negotiation (S37.0) disable HST Modulation (S13.5) enable Modem Equalization (S15.0) enable HST Mode Lower Speed (S15.2) bps450 V.21 Modulation (S27.0) disable V.32 Unencoded Modulation (S27.1) disable V.32 Modulation (S27.2) enable Bell 208 (S31.0) disable V.32 bis Modulation (S34.0) enable V.32 Enhanced Mode (S34.1) enable V.32 Fast Retrain (S34.2) enable V.23 Call Negotiation (S34.3) disable Answer Sequence (Bn) v32 Fallback Disable enable Sync Timing Source (&X) internal V.32 Terbo Modulation (S34.7) enable V.34 Modulation (S56.6) enable V.FC 2400 Symbol Rate (S54.0) enable V.FC 2743 Symbol Rate (S54.1) enable V.FC 2800 Symbol Rate (S54.2) enable V.FC 3000 Symbol Rate (S54.3) enable V.FC 3200 Symbol Rate (S54.4) enable V.FC 3429 Symbol Rate (S54.5) enable V.FC 8S-2D Mapping (S55.0) enable V.FC 16S-4D Mapping (S55.1) enable V.FC 32S-2D Mapping (S55.2) enable V.FC 64S-4D Mapping (S55.3) enable V.FC Non-linear Coding (S56.0) enable V.FC TX Level Deviation (S56.1) enable V.FC Pre-emphasis (S56.2) enable V.FC Precoding (S56.3) enable V.FC Shaping (S56.4) enable V.FC Modulation (S56.7) enable V.8 Mode (S54.7) enable V.8 Call Indicator (S54.6) disable V.34+ (S56.5) enable 300 Baud (S48.0) disable 1200 Baud (S48.1) disable 2400 Baud (S48.2) disable High Speed (S48.3) disable V.42 Selective Reject (S51.6) enable Phone Exclusion Delay (S51.7) disable Minimum High-speed Direction Link Speed (&U) variable -- Jay Nitikman (jay@cruzio.com) Cruzio is a mom and pop Internet Service Provider Web: http://www.cruzio.com Email: info@cruzio.com Voice: 423-1162
Subject: Re: (usr-tc) mdmCsStatus
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 09:53:55
At 10:33 AM 2/24/98 -0500, you wrote: >At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: >> >>I ma not sure , wether it is the same output you want, but i have to small c >>program that telnets to the netserver and does a show session and output >the stuff >>to your screen >> >>The program pwho ( shareware) is a completely c prog and easy to install .. >> >>cheers >> > >How about mentioning that it costs $500? :)) > What costs $500? Pmwho? I would like to be the one that can sell a free program for $500.. Where did this comment come from? -m
Subject: (usr-tc) Looking for used Netserver 8/I-modem
From: Netlink Support <support@netlinkcom.com>
Date: 1998-02-24 09:56:09
Does anyone out there have a used Netserver 8/I-modem for sale. (or know where to find one) I am currently using a Netserver 8/I for X2 and ISDN, but need a second one to hold me over until I can afford to move up to a PRI/CT1 solution. Thanks, Curt Schibonski Netlink Communications, Inc.
Subject: Re: (usr-tc) [Q] USR Total Security Server using 100% of CPU as Service
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 10:02:24
On Tue, 24 Feb 1998, DWX Network Operations wrote: > Greetings. > When running USR Total Security and Accounting server as a service under NT > 4.0, I find that my CPU runs right at 100% when I check the Task Manager. > Under the list of processes I see that radserver.exe fluctuates between 90 > and 95 percent of the CPU usage. Yes this is actually a problem with NT. You can open a DOS window and get the CPU to 100 %. I know that 4.x version of Radius server did show up 100 %. You have two work arounds, 1. From the task manager you can prioritize the radserve.exe process to be set as high - which helps a little. 2. Start the process as an application. I am not sure but you can try 5.x radius server - since there was ER release of security code that did lower the CPU level. However, 100% cpu does not cause any problems - nor have I heard of any krish > > When I kill the service and run the server as an application my CPU usage > is about what I would expect (8 to 10 percent). > > Is this a problem for anyone - or a documented scenario that I've > overlooked? I, of course, would prefer to run this as a service - any > words of wisdom? > > Thanks, > Tracy Hinshaw, DWX NOC > noc@dwx.com > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) unknown request type 12
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 10:10:37
On Tue, 24 Feb 1998, Edward Kern wrote: > Hi all. > > I just upgraded one of our USR TC racks to TCS 3.0.2. The NMC card, > running 5.3.2, keeps causing error messages in the syslog on my RADIUS server: > > Feb 24 18:37:00 pelican radius[5563]: unknown request type 12 from > 198.69.84.249 ignored You are getting a Status-Server code. Now who is 198.69.84.249 - is this the NETServer or the NMC? I would think that this is coming from the NETServer - the NETServer is capable of asking the status of the the server and vice versa. > > We're running BSD/OS and Livingston's RADIUS 2.0.1. The tech I talked to > at USR (we have a support contract) wouldn't help me because we're not > running USR's server..... <sigh> > This is wrong. All you are asking here is why you are getting code 12. Support may not have understood the problem properlly, let me know who it was. Also if you have opened a ticket let me know. regards krish > Anyone ever seen anything like this before? Am I overlooking something > stupid? :> > > Thanks, > > Ed. > > --- > Edward Kern (dag@soulfood.org) > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 10:30:10
I posted another message earlier regarding the bug we are experiencing here with the HiPer ARC. Basically, whenever anyone logs onto the hub with the HiPer ARC, it starts assigning random addresses to the MAC on the HiPer ARC. These addresses are usually from one of my servers which then renders the server unreachable. Usually within 20 minutes of the new hub being on-line. But, while it was on-line, our Quake players were back to the 120 pings like we had earlier. BTW, on our Total Control Hub with the Netserver card, we did not have any ping problems until we added the HiPer DSP's to the chassis. It was just too much of a load for the Netserver. Brian wrote: > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > > Unfortunately, because of some other bugs in the code, I can't use it... > > which bugs specifically? > > > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) mdmCsStatus
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-24 10:30:50
Thus spake Rick Payne >>>>>> "Brian" == Brian <signal@shreve.net> writes: > Brian> Understood. But how would I have known this? I tried to >snmpwalk is your friend. Hrmm....so I'm not the only one who used that method to find all this out then? Glad to know I'm not missing some tricks at least. :) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) mdmCsStatus
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-24 10:33:59
At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: > >I ma not sure , wether it is the same output you want, but i have to small c >program that telnets to the netserver and does a show session and output the stuff >to your screen > >The program pwho ( shareware) is a completely c prog and easy to install .. > >cheers > How about mentioning that it costs $500? :)) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) mdmCsStatus
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-24 10:33:59
At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: > >I ma not sure , wether it is the same output you want, but i have to small c >program that telnets to the netserver and does a show session and output the stuff >to your screen > >The program pwho ( shareware) is a completely c prog and easy to install .. > >cheers > How about mentioning that it costs $500? :)) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) archives
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-24 10:36:11
Thus spake Brian >Where is the USR-TC archives at, to search on older messages. There's an ftp archive at ftp.xmission.com I believe, that's what I do...just pull down all the digest packets and keep them in my own home directory so I can do searches on them (using grep of course) whenever I want. Someone was keeping an html'ized version somewhere I believe, but don't know where off the top of my head. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) Radius Packet Debug Tool
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 10:41:35
I have put together a radius packet debug tool that uses snoop or tcpdump to gather packets. These packets are then taken apart so the Attributes are easily visible.. This is handy for debugging home grown radius or possible NAS problems.. It can also verify the MD5 auth that is included in accounting packets.. If you are interested it can be found at http://coredump.ae.usr.com/raddebug There is some sample output and a basic README file there as well.. Enjoy! -m BTW: This is for UNIX.. It has been successfully compiled on Solaris 2.5.1 and RedHat Linux 4.2.. Should work for all UNIX flavors.
Subject: Re: (usr-tc) Analog Quads Don't Work in a TC Rack?
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 10:57:34
On Tue, 24 Mar 1998, Jeff Mcadams wrote: > Thus spake Mike > >At 09:53 AM 3/24/98 -0600, I wrote: > >>USR did market the TC racks with the analog modems in them. An they cant > >>be software upgraded to digital since the hardware does not have the > >>circuitry to > >>talk on the TDM bus to a PRI card. These were sold to those who intended to > >>use them > >>on analog lines.. They still use the netserver and can take analog calls.. > >>What seems to be your > >>problem with them.. If its functionality was misrepresented to you by the > > >Just to clarify.. I missed typed above.. They will work in a chassis with > >netserver > >but the netserver "CAN'T" talk to the analog cards.. You will need a > >terminal server. > > Uhm, are you *sure* about that? I'm almost positive we got some > analog-onlys talking to a netserver at one point in time. The analog-only Quad cards were never supported by the NETServer. They do not have any packet-bus circuit in them. krish > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) mdmCsStatus
From: Brian <signal@shreve.net>
Date: 1998-02-24 10:58:18
On Tue, 24 Feb 1998, Garry Shtern wrote: > At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: > > > >I ma not sure , wether it is the same output you want, but i have to small c > >program that telnets to the netserver and does a show session and output > the stuff > >to your screen > > > >The program pwho ( shareware) is a completely c prog and easy to install .. > > > >cheers > > > > How about mentioning that it costs $500? :)) Your thinking of PMMON, which is much more advanced. Brian > > Garry Shtern shterng@akula.com > Chief Network Administrator http://www.akula.com > Akula Communications Corp. tel. (212) 292-8892 > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) restricting connect rates...
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-24 10:59:17
Jay Nitikman said once upon a time: > >I'm sorry if this has been covered before but I cannot find the answer >in the archives. I want to restrict modem connections to X2 only. I'm not sure how well it will work, but I can tell you what you don't need: >Link Rate Speed Select (&N) variable >Non-ARQ Transmit Buffer Size (S15.3) bytes1500 bytes1500 >Buffer RX During MNP Negotiation (S37.0) disable Ok >HST Modulation (S13.5) enable disable >Modem Equalization (S15.0) enable >HST Mode Lower Speed (S15.2) bps450 ok >V.21 Modulation (S27.0) disable >V.32 Unencoded Modulation (S27.1) disable >V.32 Modulation (S27.2) enable >Bell 208 (S31.0) disable >V.32 bis Modulation (S34.0) enable >V.32 Enhanced Mode (S34.1) enable >V.32 Fast Retrain (S34.2) enable >V.23 Call Negotiation (S34.3) disable disable all >Answer Sequence (Bn) v32 >Fallback Disable enable >Sync Timing Source (&X) internal ok >V.32 Terbo Modulation (S34.7) enable >V.34 Modulation (S56.6) enable >V.FC 2400 Symbol Rate (S54.0) enable >V.FC 2743 Symbol Rate (S54.1) enable >V.FC 2800 Symbol Rate (S54.2) enable >V.FC 3000 Symbol Rate (S54.3) enable >V.FC 3200 Symbol Rate (S54.4) enable >V.FC 3429 Symbol Rate (S54.5) enable >V.FC 8S-2D Mapping (S55.0) enable >V.FC 16S-4D Mapping (S55.1) enable >V.FC 32S-2D Mapping (S55.2) enable >V.FC 64S-4D Mapping (S55.3) enable >V.FC Non-linear Coding (S56.0) enable >V.FC TX Level Deviation (S56.1) enable >V.FC Pre-emphasis (S56.2) enable >V.FC Precoding (S56.3) enable >V.FC Shaping (S56.4) enable >V.FC Modulation (S56.7) enable >V.8 Mode (S54.7) enable >V.8 Call Indicator (S54.6) disable >V.34+ (S56.5) enable >300 Baud (S48.0) disable >1200 Baud (S48.1) disable >2400 Baud (S48.2) disable >High Speed (S48.3) disable disable all >V.42 Selective Reject (S51.6) enable >Phone Exclusion Delay (S51.7) disable >Minimum High-speed Direction Link Speed (&U) Ok.
Subject: Re: (usr-tc) mdmCsStatus
From: Garry Shtern <shterng@akula.com>
Date: 1998-02-24 11:01:59
At 09:53 AM 2/24/98 -0600, Mike wrote: >At 10:33 AM 2/24/98 -0500, you wrote: >>At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: >>> >>>I ma not sure , wether it is the same output you want, but i have to small c >>>program that telnets to the netserver and does a show session and output >>the stuff >>>to your screen >>> >>>The program pwho ( shareware) is a completely c prog and easy to install .. >>> >>>cheers >>> >> >>How about mentioning that it costs $500? :)) >> >What costs $500? Pmwho? I would like to be the one that can sell a free >program for $500.. Where did this >comment come from? oh sorry.. I meant pmmon :))) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Jose de Leon <jadiel@thevision.net>
Date: 1998-02-24 11:02:37
Let me explain what happened since I just caught this conversation. I installed a total control chassis with netserver, pri, quad modems, nmc. This server is for dialup with dynamic IP assignment. I specified in the setup an address range from 209.60.250.130 to 209.60.250.175 for a total of 46 IPs. This total control chassis would then, at random, assign IPs outside of that range to its net0 interface. Some IPs would be 209.60.250.230, 209.60.250.34. It became quite annoying because many of these IPs were assigned to our NT servers and Win95 workstations. Needless to say, the Win95 workstations would deactivate the ether interface because it complaind that the address was in use. I discovered the interface useing the address by referencing the MAC address. So, there's my story. I couldn't give you any readouts right now since the server is now enabled with Proxy Arp and seems to work and I'm reluctant to mess with it since its loaded right now with connections. However, I'm getting another one 30 days and if I get the same problem, I'll post on the list. -----Original Message----- On Tue, 24 Feb 1998, Jose de Leon wrote: > I'm not use ARC, and I get the same problem, the technician from USR > suggested I enable proxy arp. I just don't think it should be necessary. > The interface should not assign an IP address that is not specified for the > system. > Where are you "seeing" this IP address assignment? Are you looking at the routing tables? The routing tables are *going* to show all the servers on your lan reachable via the gateway of your usr box. I am wondering if that is what you mean by "assigning the ip address of the usr box to our servers" type stuff. Can you post some screen dumps? This would be helpful. > -----Original Message----- > From: Brian <signal@shreve.net> > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > Date: Tuesday, February 24, 1998 9:38 AM > Subject: Re: (usr-tc) Quakeworld and Quake2 lag > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > I posted another message earlier regarding the bug we are experiencing > > here with the HiPer ARC. Basically, whenever anyone logs onto the hub > > with the HiPer ARC, it starts assigning random addresses to the MAC on > > the HiPer ARC. These addresses are usually from one of my servers which > > then renders the server unreachable. Usually within 20 minutes of the > > new hub being on-line. > > What version of code are you using? It must be something specific your > doing, as I have never seen this posted here before, and I know there are > alot of people using ARC's without this problem. We have run the release > code and 3 or 4 ER codes and never saw it "proxyarp" server addresses to > our MAC of our ARC. > > > > > But, while it was on-line, our Quake players were back to the 120 pings > > like we had earlier. BTW, on our Total Control Hub with the Netserver > > card, we did not have any ping problems until we added the HiPer DSP's > > to the chassis. It was just too much of a load for the Netserver. > > > > Brian > > > Brian wrote: > > > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > > > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > > > > Unfortunately, because of some other bugs in the code, I can't use > it... > > > > > > which bugs specifically? > > > > > > > > > -- > > ======================================================= > > =========== Andrew Aken - President ========= > > ====== GlobalEyes Communications, Inc. ====== > > =Southern Illinois' Fastest Connection to the Internet= > > ========== http://www.GlobalEyes.net ======== > > ======================================================= > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/ - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) mdmCsStatus
From: barberw@tidewater.net
Date: 1998-02-24 11:19:51
> This will work but it requires a SNMP get for each modem in the chassis.. > This is slow.. A better method is > to query the Chassis LED's.. This would be two gets per chassis (one for > on/off and one for color if you only want > to count the calls that are connected and "green").. The OID's for this > were in a message I sent a long time ago to the list.. Check the archives > for them.. If you cant find them I will dig them up and repost it.. When > using this on a > HDSP card you cant get an exact count of calls up, but you can get a pretty > good estimate by multiplying the number of lit LED's by 2.3 for PRI (10 > leds * 2.3 = 23 calls) or 2.4 for T1. I check for connected calls in one get. I use snmx to output to a file and was using mdmCsStatus, which was slow. Since then I've cut it down to just checking the number of interfaces on the Netserver card. I load the special USR mibs, but it isn't necessary. Here's the code: connect netserver mib -load usr.mdb cd /mgmt/mib/interfaces set -exec $busy = get ifNumber.0 set $busy = [$busy.1 - 1] set -exec $date = "date '+%H:%M %a %d/%m/%y'" echo $date.1 $busy >>modem.output ifNumber.0 is a count of current interfaces on the Netserver and one needs to be subtracted because the LAN port gets counted. If you don't want to load the usr mib, change line 3 to: cd /mgmt/mib-2/interfaces - Wayne Barber - barberw@tidewater.net Internet System Administrator Coastal Telco Services
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian <signal@shreve.net>
Date: 1998-02-24 11:27:16
On Tue, 24 Feb 1998, Andrew Aken wrote: > I posted another message earlier regarding the bug we are experiencing > here with the HiPer ARC. Basically, whenever anyone logs onto the hub > with the HiPer ARC, it starts assigning random addresses to the MAC on > the HiPer ARC. These addresses are usually from one of my servers which > then renders the server unreachable. Usually within 20 minutes of the > new hub being on-line. What version of code are you using? It must be something specific your doing, as I have never seen this posted here before, and I know there are alot of people using ARC's without this problem. We have run the release code and 3 or 4 ER codes and never saw it "proxyarp" server addresses to our MAC of our ARC. > > But, while it was on-line, our Quake players were back to the 120 pings > like we had earlier. BTW, on our Total Control Hub with the Netserver > card, we did not have any ping problems until we added the HiPer DSP's > to the chassis. It was just too much of a load for the Netserver. > Brian > Brian wrote: > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > > > Unfortunately, because of some other bugs in the code, I can't use it... > > > > which bugs specifically? > > > > > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Couriers connecting to TC at x2
From: Brian <signal@shreve.net>
Date: 1998-02-24 11:28:45
On Tue, 24 Feb 1998, chris loelke wrote: > i have been struggling with this problem since we switched to TC. > > i have had USR looking into this since january. I can call the socrates > server (usr bbs) without problems (0-30 BLERS. If i call our own TC, i > get about 1100 BLERS per minute. Sometimes (1 out of 10), i get zero BLERS. If > i get a zero BLER connection, i can stay on for hours without a problem.... > > The other thing i noticed was the the SNR (signal to noise ratio) is around > 36 db when i get BLERS, but around 50 db when i get zero BLERS. > > > This same modem, when used at a different exchange > has no problems at all. Some of our other users mainly X2, are having the > same problem, but they complain that it is slow to use the system. > > > This issue has been escalated to level 2 at USR, i am trying to escalate this > to level 999999. Sounds more like a telco problem. Sounds like maybe switch pads or something. If you can dial long distance and x2 or whatever is better, than a intraLATA call, then its probably pads. Brian > > > At 10:41 AM 2/23/98 -0600, you wrote: > >I have a Courier modem with x2. Recently, I started seeing 1000s of BLERs > >when connected to a TC rack with modem code 5.6.7. > > > >I thought the modem was bad, so I swapped it for another Courier running > >the 02/28/97 code. The errors went away. Then I upgraded the code to the > >7/97 code in the Courier. I started to get 1000s of errors again. > >Downgraded to the 2/97 code, and it all works again. > > > >Any ideas if this is a Courier issue, or a TC rack issue? > > > >I haven't tried upgrading the modem code on the TC rack. Should I do > >that? I was just going to wait for the v.90 code for my next upgrade, but > >maybe I shouldn't wait. > > > >Brian > > > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver idle-timeout
From: jason_kelton@3com.com
Date: 1998-02-24 11:30:53
Richard, You might want to try upgrading to NETServer 3.7.24... and while you're at it, you might need to upgrade your NETS with 16 Meg of RAM. Regards, Jason. bosire@ns1.africaonline.co.ke on 23/02/98 19:51:23 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) Hi .. I have this problem of netserver idle-timeout ,, The suppossedly working feature does not work on my setup .. Is anyone having the same problem , is there any work-around .. my netserver details are: U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.3.28 Build date: Dec 13 1996 Build time: 13:54:59 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Enhanced Licensed for 60 ports. ] cheers and thanx -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/ - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) mdmCsStatus
From: Andy Scutt <andy.scutt@ntli.net>
Date: 1998-02-24 11:36:10
Hiya We use MRTG (Multi-router Traffic Grapher) information of which is at: http://ee-staff.ethz.ch/~oetiker/webtools/mrtg/mrtg.html Give it a go, I think its what you are looking for. Andy Scutt Network Engineer NTL Internet andy.scutt@ntli.net -----Original Message----- >I am writing a program so I can tell how many users are logged on all my >hubs at any given time, if nothing else, I am also trying to get an >education in ucd SNMP. > >I am querying mdmCsStatus to find out if the modem is in use, let me know >if there is a better way. > >My program is not working, here is what I have: > > >use SNMP; >SNMP::setMib("/usr/local/mibs/mib.txt"); > >$host = shift; >unless ($host) { > $| = 1; print "enter SNMP host address: "; $| = 0; > chomp($host = <STDIN>); >} > >$secret = shift; >unless ($secret) { > $| = 1; print "enter SNMP community string: "; $| = 0; > chomp($secret = <STDIN>); >} > >$session = new SNMP::Session ( DestHost => $host, Community => $secret) || die; > >for ($i=0; $i<48; $i++) { > $mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > $session->get($mdmCsStatus); > print "> $mdmCsStatus\n"; >} > >I think my problem has to do with this line: > >$mdmCsStatus = new SNMP::Varbind(['mdmCsStatus', $i]); > >Does anyone know what I could be doing wrong? Thanks. > >Also if anyone has already patched pmwho to give some kind of useful >output when used with a ARC that would be great and save me some time, >otherwise thats my next project. > >Brian > > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Network Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) TC Hub HiPer Arc IP allocation problems
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-24 11:37:41
Andrew Aken said once upon a time: > >We just set up a new TC hub with the HiPer Arc router card. We currently >are using some RADIUS servers to do our authentication (AimTraveler >RADIUS running on NT). Everything looks fine for a while, but then the >HiPer Arc card starts stealing IP addresses from the rest of our >network. We've verified this by matching the MAC address of the HiPer >Arc with the hardware address in the NT log files. The RADIUS log files, >however, show that the IP addresses given to the callers are from the IP >Address pool in the HiPer Arc configuration. > >Does anyone understand what I am talking about (it's very late and I've >been working on this for 3 straight days)? Does anyone have any clues on >where to fix this bug? I really need to get the new hub up and running >ASAP. What does "show networks" show for your callers? I don't see how it could be "stealing addresses" unless what you mean it is advertising addresses, in which case, it sounds like you need split-horizon and/or poison-reverse turned on.
Subject: Re: (usr-tc) mdmCsStatus
From: James Wilson <james@cruxnet.com>
Date: 1998-02-24 11:51:03
I wish I cought this thread earlier. This is how we used to do it on our Portmasters, and how we do it now on our TC Hubs. James > From: barberw@tidewater.net > >> This will work but it requires a SNMP get for each modem in the chassis.. >> This is slow.. A better method is >> to query the Chassis LED's.. This would be two gets per chassis (one for >> on/off and one for color if you only want >> to count the calls that are connected and "green").. The OID's for this >> were in a message I sent a long time ago to the list.. Check the archives >> for them.. If you cant find them I will dig them up and repost it.. When >> using this on a >> HDSP card you cant get an exact count of calls up, but you can get a pretty >> good estimate by multiplying the number of lit LED's by 2.3 for PRI (10 >> leds * 2.3 = 23 calls) or 2.4 for T1. > >I check for connected calls in one get. > >I use snmx to output to a file and was using mdmCsStatus, which was >slow. Since then I've cut it down to just checking the number of >interfaces on the Netserver card. I load the special USR mibs, but it >isn't necessary. Here's the code: > >connect netserver >mib -load usr.mdb >cd /mgmt/mib/interfaces >set -exec $busy = get ifNumber.0 >set $busy = [$busy.1 - 1] >set -exec $date = "date '+%H:%M %a %d/%m/%y'" >echo $date.1 $busy >>modem.output > >ifNumber.0 is a count of current interfaces on the Netserver and one >needs to be subtracted because the LAN port gets counted. > >If you don't want to load the usr mib, change line 3 to: >cd /mgmt/mib-2/interfaces > > >- >Wayne Barber - barberw@tidewater.net >Internet System Administrator >Coastal Telco Services
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian <signal@shreve.net>
Date: 1998-02-24 11:52:09
On Tue, 24 Feb 1998, Jose de Leon wrote: > I'm not use ARC, and I get the same problem, the technician from USR > suggested I enable proxy arp. I just don't think it should be necessary. > The interface should not assign an IP address that is not specified for the > system. > Where are you "seeing" this IP address assignment? Are you looking at the routing tables? The routing tables are *going* to show all the servers on your lan reachable via the gateway of your usr box. I am wondering if that is what you mean by "assigning the ip address of the usr box to our servers" type stuff. Can you post some screen dumps? This would be helpful. > -----Original Message----- > From: Brian <signal@shreve.net> > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > Date: Tuesday, February 24, 1998 9:38 AM > Subject: Re: (usr-tc) Quakeworld and Quake2 lag > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > I posted another message earlier regarding the bug we are experiencing > > here with the HiPer ARC. Basically, whenever anyone logs onto the hub > > with the HiPer ARC, it starts assigning random addresses to the MAC on > > the HiPer ARC. These addresses are usually from one of my servers which > > then renders the server unreachable. Usually within 20 minutes of the > > new hub being on-line. > > What version of code are you using? It must be something specific your > doing, as I have never seen this posted here before, and I know there are > alot of people using ARC's without this problem. We have run the release > code and 3 or 4 ER codes and never saw it "proxyarp" server addresses to > our MAC of our ARC. > > > > > But, while it was on-line, our Quake players were back to the 120 pings > > like we had earlier. BTW, on our Total Control Hub with the Netserver > > card, we did not have any ping problems until we added the HiPer DSP's > > to the chassis. It was just too much of a load for the Netserver. > > > > Brian > > > Brian wrote: > > > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > > > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > > > > Unfortunately, because of some other bugs in the code, I can't use > it... > > > > > > which bugs specifically? > > > > > > > > > -- > > ======================================================= > > =========== Andrew Aken - President ========= > > ====== GlobalEyes Communications, Inc. ====== > > =Southern Illinois' Fastest Connection to the Internet= > > ========== http://www.GlobalEyes.net ======== > > ======================================================= > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Netserver idle-timeout
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-24 11:54:43
Richard Bosire said once upon a time: >Hi Pete ,No , I have been trying it from the netserver command prompt .. >Pliz you dont mind shedding some ideas on how to do this from RADIUS I just have the following line in my DEFAULT user entry: Idle-Timeout = 600
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-24 12:12:29
Brian said once upon a time: >I thought Sel Rej *helped* with quake.........your saying you disable >selrej, to help quake? Right. That was Krish's suggestion a few months ago.
Subject: Re: (usr-tc) IP Address Conflicts
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 12:33:07
From my NT error logs I get messages like: The system detected an address conflict for IP address 209.60.64.5 with the system having network hardware address 00:C0:49:0D:5F:37. The local interface has been disabled. The network hardware address is the MAC of eth:1 on the HiPer ARC. Of course, since we're running an IP network, when the local interface to IP is disabled, it renders the machine unreachable from anywhere else on the network. Brian wrote: > > Where are you "seeing" this IP address assignment? Are you looking at the > routing tables? The routing tables are *going* to show all the servers on > your lan reachable via the gateway of your usr box. I am wondering if > that is what you mean by "assigning the ip address of the usr box to our > servers" type stuff. > > Can you post some screen dumps? This would be helpful. > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-24 12:45:51
Andrew Aken said once upon a time: > >BTW, USR told me that you need to set the IP pools to 15% more IP >addresses than the system needs to service the modems in your hub. >Something to do with the time needed to release the IP addresses back >into the pool upon disconnection. Does this apply to the ARC as well?
Subject: Re: (usr-tc) restricting connect rates...
From: Brian <signal@shreve.net>
Date: 1998-02-24 13:02:36
> >V.8 Mode (S54.7) enable > >V.8 Call Indicator (S54.6) disable you sure v.8 isn't needed? Brian > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) IP Address Conflicts
From: Brian <signal@shreve.net>
Date: 1998-02-24 13:04:30
On Tue, 24 Feb 1998, Andrew Aken wrote: > >From my NT error logs I get messages like: > The system detected an address conflict for IP > address 209.60.64.5 with the system having network > hardware address 00:C0:49:0D:5F:37. The local > interface has been disabled. > The network hardware address is the MAC of eth:1 on the HiPer ARC. Of > course, since we're running an IP network, when the local interface to > IP is disabled, it renders the machine unreachable from anywhere else on > the network. have you verified all this by looking at the ARP tables on some of your servers. > > Brian wrote: > > > > Where are you "seeing" this IP address assignment? Are you looking at the > > routing tables? The routing tables are *going* to show all the servers on > > your lan reachable via the gateway of your usr box. I am wondering if > > that is what you mean by "assigning the ip address of the usr box to our > > servers" type stuff. > > > > Can you post some screen dumps? This would be helpful. > > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 13:21:02
At 11:02 AM 2/24/98 -0800, you wrote: >Let me explain what happened since I just caught this conversation. > >I installed a total control chassis with netserver, pri, quad modems, nmc. >This server is for dialup with dynamic IP assignment. I specified in the >setup an address range from 209.60.250.130 to 209.60.250.175 for a total of >46 IPs. This total control chassis would then, at random, assign IPs >outside of that range to its net0 interface. Some IPs would be >209.60.250.230, 209.60.250.34. It became quite annoying because many of >these IPs were assigned to our NT servers and Win95 workstations. Needless >to say, the Win95 workstations would deactivate the ether interface because >it complaind that the address was in use. I discovered the interface useing >the address by referencing the MAC address. > >So, there's my story. I couldn't give you any readouts right now since the >server is now enabled with Proxy Arp and seems to work and I'm reluctant to >mess with it since its loaded right now with connections. However, I'm >getting another one 30 days and if I get the same problem, I'll post on the >list. After you set up the pools and the limit did you save all & reboot? The ip pool and limit only take effect after a reboot.. This has caused many people some grief.. -m
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 13:31:35
So is there a comparable setting to Proxy ARP on the HiPer ARC? Jose de Leon wrote: > > > So, there's my story. I couldn't give you any readouts right now since the > server is now enabled with Proxy Arp and seems to work and I'm reluctant to > mess with it since its loaded right now with connections. However, I'm > getting another one 30 days and if I get the same problem, I'll post on the > list. > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 13:34:43
BTW, USR told me that you need to set the IP pools to 15% more IP addresses than the system needs to service the modems in your hub. Something to do with the time needed to release the IP addresses back into the pool upon disconnection. Mike wrote: > > After you set up the pools and the limit did you save all & reboot? The ip > pool and limit only take effect after > a reboot.. This has caused many people some grief.. > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 13:47:47
At 01:34 PM 2/24/98 -0600, you wrote: >BTW, USR told me that you need to set the IP pools to 15% more IP >addresses than the system needs to service the modems in your hub. >Something to do with the time needed to release the IP addresses back >into the pool upon disconnection. This is only true if you are using "hint assigned" or PPTP. If you are not using these set your limit to the correct value. >Mike wrote: >> >> After you set up the pools and the limit did you save all & reboot? The ip >> pool and limit only take effect after >> a reboot.. This has caused many people some grief.. >> `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)
Subject: Re: (usr-tc) restricting connect rates...
From: Russ Panula <rpanula@dacmail.net>
Date: 1998-02-24 14:12:20
On Tue, 24 Feb 1998 09:45:38 -0800, Jay Nitikman <jay@cruzio.com> wrote: >I'm sorry if this has been covered before but I cannot find the answer >in the archives. I want to restrict modem connections to X2 only. [snip] >Minimum High-speed Direction Link Speed (&U) variable Set this to the lowest X2 speed (33.333bps). That alone should limit the connections to X2...
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 14:17:31
What is "hint assigned" Tatai SV Krishnan wrote: > > It is only true if you are using Hint assigned - If you are using PPTP > then the address is given out by NT RAS so you need to setup the pool on > NT not on NETServer. > > If you are using Hint assigned on the Hiper ARC we do suggest to increase > your pool by 2 or more addresses. > > krish > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) mdmCsStatus
From: Rick Payne <rickp@corp.netcom.net.uk>
Date: 1998-02-24 14:49:06
>>>>> "Brian" == Brian <signal@shreve.net> writes: Brian> Understood. But how would I have known this? I tried to snmpwalk is your friend. Rick -- Rick Payne, Senior Network Admin | Taxation is about extracting the NETCOM Internet Ltd. | maximum amount of milk with the rickp@corp.netcom.net.uk | minimum of moo.
Subject: Re: (usr-tc) Couriers connecting to TC at x2
From: Brian Elfert <brian@citilink.com>
Date: 1998-02-24 14:57:56
On Tue, 24 Feb 1998, Brian wrote: > On Tue, 24 Feb 1998, chris loelke wrote: > > > i have been struggling with this problem since we switched to TC. > > > > i have had USR looking into this since january. I can call the socrates > > server (usr bbs) without problems (0-30 BLERS. If i call our own TC, i > > get about 1100 BLERS per minute. Sometimes (1 out of 10), i get zero BLERS. If > > i get a zero BLER connection, i can stay on for hours without a problem.... > > > > The other thing i noticed was the the SNR (signal to noise ratio) is around > > 36 db when i get BLERS, but around 50 db when i get zero BLERS. Upgrade to the latest code on your TC's modems. This fixed my problem with 1000s of Blers with a Courier modem. Brian
Subject: (usr-tc) Bug Reports and tech support
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-24 15:06:52
Just spoke to Skip Lasher at 3Com...he had gotten a copy of my previous email about putting in bug reports... We talked about the bug reporting process a bit and I wanted to report the results of that conversation out.... He said that the original tech that I spoke to was in San Diego which is not equipped for handling all the equipment that they deal with so if your call goes through there, they might end up putting you back into the hold queue to be picked up in their Chicago facility (not so bad of a situation now that hold times are down pretty short) where they are better equipped to handle questions about some equipment. In Chicago, they should be able to determine on their own whether something you are reporting should be reported to Engineering or whatever, though my experience (admittedly limited) showed that a little prompting might not hurt. :) San Diego was apparently picked up in the 3Com merger, so they aren't quite up to speed on all the USR equipment yet from what I could tell, but they do seem to be getting a bit better...we'll see. Overall, my impression is the tech support is heading in the right direction, but still has a ways to go. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) IP Pool BS [Was: Quakeworld and Quake2 lag]
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-24 15:30:46
Andrew Aken was heard to say: >BTW, USR told me that you need to set the IP pools to 15% more IP >addresses than the system needs to service the modems in your hub. >Something to do with the time needed to release the IP addresses back >into the pool upon disconnection. What the f*ck is this BS? The instant the PPP/SLIP/whatever session is dropped, the address should be returned to the free address pool. What kind of stupid programming BS have they done? What did they, write the NAS firmware in JAVA? --Ricky
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-24 15:37:13
Are there any settings on the TCH which I must make to use "hint assigned" addressing? Tatai SV Krishnan wrote: > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > What is "hint assigned" > > > > Hint assigned is a way of telling the radius server what IP to give when > a user dials in. This is something that you have to do by modifying the > radius server. What happens is when a user comes in, and is > authenticated by Radius, the NETServer/HiPer arc can send the Radius and > IP address from its pool that will be assigned to the user. This method > is called hint assigned. > > krish > > > ======================================================= > > =========== Andrew Aken - President ========= > > ====== GlobalEyes Communications, Inc. ====== > > =Southern Illinois' Fastest Connection to the Internet= > > ========== http://www.GlobalEyes.net ======== > > ======================================================= > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: Re: (usr-tc) Radius Attributes...
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-24 15:59:07
At 04:12 PM 2/24/98 -0500, you wrote: > >I Don't happen to have the USR Tc books with me at the present time (At >home), does anyone remember what attribute 25,32 are? >Also, if anyone remembers the TC radius PDF that lists the dictionary i >would be most app. > 25 = Class 32 = NAS-Identifier See RFC for more details.. -m
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian <signal@shreve.net>
Date: 1998-02-24 16:12:35
On Tue, 24 Feb 1998, Jose de Leon wrote: > Let me explain what happened since I just caught this conversation. > > I installed a total control chassis with netserver, pri, quad modems, nmc. > This server is for dialup with dynamic IP assignment. I specified in the > setup an address range from 209.60.250.130 to 209.60.250.175 for a total of > 46 IPs. This total control chassis would then, at random, assign IPs > outside of that range to its net0 interface. Some IPs would be > 209.60.250.230, 209.60.250.34. It became quite annoying because many of > these IPs were assigned to our NT servers and Win95 workstations. Needless > to say, the Win95 workstations would deactivate the ether interface because > it complaind that the address was in use. I discovered the interface useing > the address by referencing the MAC address. so you did this with what? like this? set assigned 209.60.250.130 set limit 46 so you have this netserver, with a dialup pool of: 209.60.250.130 - 209.60.250.175 which totally falls on *no* subnetting boundry whatsoever, not that it matters, but would make things more clean. Now you have machines which are assigned both above and below that pool, and they are on your ethernet.............is your netserver on that ethernet also? I don't like the idea of the ip pool, net0, and servers all on the same network. Brian > > So, there's my story. I couldn't give you any readouts right now since the > server is now enabled with Proxy Arp and seems to work and I'm reluctant to > mess with it since its loaded right now with connections. However, I'm > getting another one 30 days and if I get the same problem, I'll post on the > list. > > > -----Original Message----- > From: Brian <signal@shreve.net> > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > Date: Tuesday, February 24, 1998 9:57 AM > Subject: Re: (usr-tc) Quakeworld and Quake2 lag > > > On Tue, 24 Feb 1998, Jose de Leon wrote: > > > I'm not use ARC, and I get the same problem, the technician from USR > > suggested I enable proxy arp. I just don't think it should be necessary. > > The interface should not assign an IP address that is not specified for > the > > system. > > > > Where are you "seeing" this IP address assignment? Are you looking at the > routing tables? The routing tables are *going* to show all the servers on > your lan reachable via the gateway of your usr box. I am wondering if > that is what you mean by "assigning the ip address of the usr box to our > servers" type stuff. > > Can you post some screen dumps? This would be helpful. > > > > -----Original Message----- > > From: Brian <signal@shreve.net> > > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > > Date: Tuesday, February 24, 1998 9:38 AM > > Subject: Re: (usr-tc) Quakeworld and Quake2 lag > > > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > I posted another message earlier regarding the bug we are experiencing > > > here with the HiPer ARC. Basically, whenever anyone logs onto the hub > > > with the HiPer ARC, it starts assigning random addresses to the MAC on > > > the HiPer ARC. These addresses are usually from one of my servers which > > > then renders the server unreachable. Usually within 20 minutes of the > > > new hub being on-line. > > > > What version of code are you using? It must be something specific your > > doing, as I have never seen this posted here before, and I know there are > > alot of people using ARC's without this problem. We have run the release > > code and 3 or 4 ER codes and never saw it "proxyarp" server addresses to > > our MAC of our ARC. > > > > > > > > But, while it was on-line, our Quake players were back to the 120 pings > > > like we had earlier. BTW, on our Total Control Hub with the Netserver > > > card, we did not have any ping problems until we added the HiPer DSP's > > > to the chassis. It was just too much of a load for the Netserver. > > > > > > > Brian > > > > > Brian wrote: > > > > > > > > On Tue, 24 Feb 1998, Andrew Aken wrote: > > > > > > > > > FWIW, the new HiPer ARC card fixed the Quake lag problem for me. > > > > > Unfortunately, because of some other bugs in the code, I can't use > > it... > > > > > > > > which bugs specifically? > > > > > > > > > > > > -- > > > ======================================================= > > > =========== Andrew Aken - President ========= > > > ====== GlobalEyes Communications, Inc. ====== > > > =Southern Illinois' Fastest Connection to the Internet= > > > ========== http://www.GlobalEyes.net ======== > > > ======================================================= > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the message. > > > For information on digests or retrieving files and old messages send > > > "help" to the same address. Do not use quotes in your message. > > > > > > > /-------------------------- > signal@shreve.net -----------------------------\ > > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 > | > > | Network Administrator | Perl, Linux | Web hosting, online stores, > | > > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs > | > > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ > | > > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 > | > > \-------------------------- 318-222-2638 > x109 -----------------------------/ > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: (usr-tc) Radius Attributes...
From: Scott Portmaster List <scottport@mail.rpa.net>
Date: 1998-02-24 16:12:56
I Don't happen to have the USR Tc books with me at the present time (At home), does anyone remember what attribute 25,32 are? Also, if anyone remembers the TC radius PDF that lists the dictionary i would be most app. Thanks, Scott T. Traynor
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Brian <signal@shreve.net>
Date: 1998-02-24 16:13:49
On Tue, 24 Feb 1998, Andrew Aken wrote: > BTW, USR told me that you need to set the IP pools to 15% more IP > addresses than the system needs to service the modems in your hub. > Something to do with the time needed to release the IP addresses back > into the pool upon disconnection. I can imagine this. Can someone confirm this? I would imagine if you were in 100% usage this would be a trouble. I just put /26's on the 48 port boxes. Brian > > Mike wrote: > > > > After you set up the pools and the limit did you save all & reboot? The ip > > pool and limit only take effect after > > a reboot.. This has caused many people some grief.. > > > -- > ======================================================= > =========== Andrew Aken - President ========= > ====== GlobalEyes Communications, Inc. ====== > =Southern Illinois' Fastest Connection to the Internet= > ========== http://www.GlobalEyes.net ======== > ======================================================= > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Bug Reports and tech support
From: Brian <signal@shreve.net>
Date: 1998-02-24 16:18:22
On Tue, 24 Feb 1998, Jeff Mcadams wrote: > Just spoke to Skip Lasher at 3Com...he had gotten a copy of my previous > email about putting in bug reports... > Ive spoken to skip myself, he is a really nice guy, genuinely concerned with the service they provide, I feel comfortable things will go in a good direction with him at the helm. Brian > -- > Jeff McAdams Email: jeffm@iglou.com > Chief Network Administrator Voice: (502) 966-3848 > IgLou Internet Services (800) 436-4456 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) HELP! Authentication Issues with HiPerARC
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 16:22:37
Do an _auth username password from the ARC. Where the username is the radius username and the password is his password. This will tell if the user name is authenticated or not. Based on that information we can see if the HiPer ARC is sending request or not and if it is sending the info we can find out to which Radius server. krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec On Mon, 23 Mar 1998, Marshall Morgan wrote: > Ok ... so I am no dummy when it comes to Radius Authentication and 3COM/USR TC > products. I read the list, get several ARC's in, play with them for a few days > and then ... put them in service - (2) ARCs and (6) HiperDSPs. Flawless > install. 10 days later all is well, so we decided to fire some more up. Go > over every detail to ensure a smooth cutover, busy out the old chassis' and > when everyone is gone, boom! ... we make the cut. Phew .... all done. Success > Success! <Queen's "We are the Champions" plays in the background> > > A few minutes later the first of many calls to these shiny new chassis' arrive. > Not authenticated. Checked the authentication parameters, all looks fine. > Cut and paste the secret and such again to ensure this is right. Still no go. > So we jump over to the other arc in this chassis and same thing. Callers are > not getting authenticated. Just to be sure I understand what is going on we > get on the radius server and kill and restart radius in debug mode ... no pa > ckets from the newly installed ARC's! What?? > > hiper2-arc2> sh authentication > > AUTHENTICATION SETTINGS > Local Authentication is: ENABLED > RADIUS Authentication is: ENABLED > Hint Assigned is: DISABLED > Primary Server is: 208.137.128.29 > Primary Server Port is: 1645 > Secondary Server is: 208.137.128.7 > Secondary Server Port is: 1645 > Retransmission Timeout: 3 > Max Retranmissions: 10 > > hiper2-arc2> sh authentication counters > > AUTHENTICATION COUNTERS > Local Successful Authentications: 1 > Local Failed Authentications: 0 > Remote Successful Authentications: 0 > Remote Failed Authentications: 0 > Remote No Responses: 0 > > ============== > > What is going on ?? What to do? I have checked and checked and 2 people went > behind me and have done the same thing. Any ideas? Rad log files are not > unhappy and the /etc/raddb/clients files are correct. ARG!! > > Running V4.0.69 of the ARC code. > > Marshall Morgan > Confused! > > Internet Doorway, Inc. (aka NETDOOR) > http://www.netdoor.com > 601.969.1434 Ext. #28 | Fax 601.969.3838 | 800.952.1570 Ext. #28 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Sample quad configuration..
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-24 16:30:41
System Administrator said once upon a time: > >Can someone post a sample quad configuration that will accept ISDN calls? >I tried setting my ISDN call control options to the *exact* same thing as >my HDM ISDN call control, and now ISDN can't connect to the quads (yes, >I've got ISDN-GW set to 0). I'm trying to get better performance from the >quads than I do from the Munich, and not having much luck. What version are your Quads on, and what are those settings specifically?
Subject: (usr-tc) Sample quad configuration..
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-24 18:01:35
Can someone post a sample quad configuration that will accept ISDN calls? I tried setting my ISDN call control options to the *exact* same thing as my HDM ISDN call control, and now ISDN can't connect to the quads (yes, I've got ISDN-GW set to 0). I'm trying to get better performance from the quads than I do from the Munich, and not having much luck. Many thanks, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Bug Reports and tech support
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-24 18:11:38
Thus spake Brian >On Tue, 24 Feb 1998, Jeff Mcadams wrote: >> Just spoke to Skip Lasher at 3Com...he had gotten a copy of my previous >> email about putting in bug reports... >Ive spoken to skip myself, he is a really nice guy, genuinely concerned >with the service they provide, I feel comfortable things will go in a good >direction with him at the helm. Indeed...I must agree with your assessment of him (since many messages here tend to make it to the people that we're talking about, I figure its good to pay compliments here :), and I have noticed that tech support is moving in the right direction...I have had a remarkable streak of people recently actually dealing with problems that demonstrated a decent clue level too...it really is getting better, but it really did have a long way to go too. Now if the cost of support contracts would just get dropped to something reasonble, but I don't believe that's really within Skip's purview. For the price of a 3Com support contract, Cisco would most likely have people on site within 24 hours if anything went wrong (for comparable equipment of course). -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 18:35:21
jeff.binkley@asacomp.com (Jeff Binkley) writes: > Is there a target number of modems per query you use to avoid exceeding > the PDU length limit of the NMC or possibly a fragmentation problem ? > I've not looked at the length of an individual modem query PDU nor the > reply packet. Yes, I use a highly scientific, arduously researched, empirically tested algorithm for judging the objects to query simultaneously. 17. :-) Actually, that's the blocking size I first started with in '94 and while some apps tweak it for specific known queries, it's still the general default I use - by and large it has proven to be a reasonable trade-off between packet size versus performance while not exceeding NMC limitations. And so far it's worked with just about any combination of objects in the tree in a single query. It'll blow up if you try to query lots of long string objects simultaneously but I haven't found a table in the MIB that is likely to create that condition for a single query - at least not yet :-) Originally, I estimated that based on a full chassis query of card objects (which themselves are 12-element OIDs) assuming an average depth of a USR OID, and the desire to be able to query complete slot information for an entire chassis in a single PDU. The original NMCs couldn't fragment or handle responses more than around 576 bytes. I thought that changed at some point, but now I'm not quite so sure since a quick query to a latest version NMC still seems to have a problem with very large responses. In any event, there's a potential negative to really large queries anyway, since you then have to allow for longer timeouts for the NMC to satisfy the entire request, which for stuff like DS0 queries can be a killer. However, you can actually cram quite a few more modems into a single PDU if you are just querying this object. The actual OID in query and response is going to be 14 elements long, which as an integer, makes the query actually slightly longer than the response. Each modem uses about 23 bytes in the query, with a fixed overhead on the PDU of about 25 bytes (depending on community string, etc..). Sticking to around the 576 byte mark lets you get 23 modems in a query (darn, not quite 24). There's a high overhead cost per query to the NMC in general, so combining objects in a PDU can have a large benefit, particularly if the NMC is servicing other queries or doing other stuff like on a live rack. For some objects (particularly those serviced by older Dual-T1 cards) the intra-chassis communication can take over, but modems are pretty fast in that respect. For example, on an active rack, polled from a host 100ms away: * Walking one object at a time (48 modems): 11.1s (.23s/query) * Walking 17 objects at a time (48 modems): 2.4s (.8s/query) So as you can see there is some more overhead per individual query in the parallel case, but the overall time is vastly better, but the ratio by which overall wall clock time improves is much more than by how much the per-query time grows. This was with 486 NMCs, the difference was even more pronounced with 386 NMCs if I recall. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 18:41:42
Brian <signal@shreve.net> writes: > Understood. But how would I have known this? I tried to read the MIB and > make sense of it, and read thru the Parameter Reference. Take someone > like me, who wants to embark on writing some SNMP stuff for the TC, how do > I figure out that the modem entities are (<slot>*1000)+unit? Is this > documented somewhere, if so I will read it............I think thats my > first problem is not having all the information I need at hand. Best starting point is the NMC SNMP MIB Reference Manual. (Not the parameter reference - that's nothing more than a dump of the MIB to use up trees - or CD space). The document was released as part of NMC 3.x, but I think it's still around. It has a nice overview of how the MIBs are structured and indexed (in the "Chassis MIB" chapter). However, the MIBs also have the same information. In almost all cases in the MIB where an entity index is used, you'll find it documented as being the chassis MIB entity. For example, the index of the modem statistics table (mdmCsIndex) has a description of: "The index value which uniquely identifies the modem to which this entry is applicable. The value corresponds to the index of the same modem in the entity table of the chassis MIB." They all reference back to the chassis MIB (chs.mib) in the entity table (uchasEntityTable) where the index (uchasEntityIndex) is defined as: "A unique index for the entity for which this entry contains information. For simplicity's sake, entity index values will be assigned according to the formula E=1000*S+I, where E is the entity index, S is the slot index, and I is the index of the entity on the specific card. For example, the third modem on a quad modem card that is in slot 5 has an entity index of 5003." The primary exception to this rule is the chassis slot table (uchasSlotTable) but it's index (uchasSlotIndex) documents that it just uses slot numbers (or 20+slot number for NICs). It can be a lot of information at first, but for all their warts (and there have been plenty over time), I've generally been fairly impressed with the USR TC MIBs. They tend to be consistent, obey common rules (like the command tables) and evolve rationally over time. (Always with the occasional exception). -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: (usr-tc) unknown request type 12
From: Edward Kern <dag@soulfood.org>
Date: 1998-02-24 19:17:40
Hi all. I just upgraded one of our USR TC racks to TCS 3.0.2. The NMC card, running 5.3.2, keeps causing error messages in the syslog on my RADIUS server: Feb 24 18:37:00 pelican radius[5563]: unknown request type 12 from 198.69.84.249 ignored We're running BSD/OS and Livingston's RADIUS 2.0.1. The tech I talked to at USR (we have a support contract) wouldn't help me because we're not running USR's server..... <sigh> Anyone ever seen anything like this before? Am I overlooking something stupid? :> Thanks, Ed. --- Edward Kern (dag@soulfood.org)
Subject: Re: (usr-tc) mdmCsStatus
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-24 19:20:53
Hi .. Wel i dind'nt pay a cent for it .... Garry Shtern wrote: > At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: > > > >I ma not sure , wether it is the same output you want, but i have to small c > >program that telnets to the netserver and does a show session and output > the stuff > >to your screen > > > >The program pwho ( shareware) is a completely c prog and easy to install .. > > > >cheers > > > > How about mentioning that it costs $500? :)) > > Garry Shtern shterng@akula.com > Chief Network Administrator http://www.akula.com > Akula Communications Corp. tel. (212) 292-8892 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) mdmCsStatus
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-24 19:20:53
Hi .. Wel i dind'nt pay a cent for it .... Garry Shtern wrote: > At 08:30 AM 2/24/98 +0300, Richard Bosire wrote: > > > >I ma not sure , wether it is the same output you want, but i have to small c > >program that telnets to the netserver and does a show session and output > the stuff > >to your screen > > > >The program pwho ( shareware) is a completely c prog and easy to install .. > > > >cheers > > > > How about mentioning that it costs $500? :)) > > Garry Shtern shterng@akula.com > Chief Network Administrator http://www.akula.com > Akula Communications Corp. tel. (212) 292-8892 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 19:42:12
Mike <mwronski@coredump.ae.usr.com> writes: > This will work but it requires a SNMP get for each modem in the chassis.. > This is slow.. A better method is > to query the Chassis LED's.. The only wrinkle to this is that you have to know the contents of the chassis in order to properly apply the LED information, whereas walking the mdmCsTable doesn't require that. Of course, to get good speed out of the mdmCsTable by grabbing multiple objects you need to know contents again so I guess it's probably a wash. > When using this on a HDSP card you cant get an exact count of calls > up, but you can get a pretty good estimate by multiplying the number > of leit LED's by 2.3 for PRI (10 leds * 2.3 = 23 calls) or 2.4 for T1. (or 3.0 for E1 - PRI or channelized) FYI, for those who may be as crazy as me, I optimize this estimate for HDM cards slightly, because of the behavior that the first LED actually turns on with the very first modem (and not after 2-3) and the last LED only turns on when _all_ the modems are in use. So while I treat most active LEDs as 10% of the total available modems (according to card and circuit type), the uppermost active LED (as long as it isn't one of the top 2 LEDs of the card) is considered to only count half as much of the standard per-LED value, or 5% of total. This is to allow for the fact that the uppermost LED turned on at the bottom of the 10% threshold but might not include the full set of channels. The uppermost two LEDs are excluded since the top LED only comes on when full, and because of that the "gap" from the second to top LED is pretty much two full LED values, so when the second to top LED is on, guessing at half of that two full steps (which is really just a single normal step) is a good estimate. Get that? :-) Oh, and since this general subject seems to be getting revisited, I've also sent some rather lengthy notes about user scans to the list in the past (the archive, BTW, since someone else asked about it, at least that I use is at http://usr-tc.datasys.net/). But just to simplify things and put them in one place, I'll include two relevant older notes below (the first of which actually included an even older note below it). The first kind of covers the subject in general whereas the second has some info on the LED objects. Note that in the first note, I make mention to querying the ifNumber object for the NETServer as another message has indicated. That is definitely a possibility, but as pointed out below does have some disadvantages depending on your user base and just what you are trying to query. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/ - - - - - - - - - - - - - - - - - - - - - - - - - In-Reply-To: Your message of Mon, 12 Jan 1998 14:26:52 +0100 Message-ID: <CMM.0.90.2.884635228.db3l@valheru.ny.ans.net> Robert von Bismarck <rvb@petrel.ch> writes: > This is probably documented somewhere, but I need to know how many users > are currently connected to a TC Hub via SNMP (so I can list this in my > MRTG stats). There are several ways, depending on which card(s) you wish to query for that information. However, each method also has some pros and cons associated with it. You can't get it from the NETServer via SNMP since it only has MIB-II and while you could check interfaces for PPP customers or TCP streams for non-PPP, it wouldn't include users still logging on. You can, however, probably get more of what you want with an embedded telnet session to do a "show all" or "show session". If all you service is PPP customers, then the other response posted her that checks for the interfaces should work pretty well, although it'll only include users who have successfully started the PPP negotiation such that the NETServer has added an interface for them. As far as other SNMP options, there are probably three possible methods: (1) Querying the DS0 status on any line cards in the chassis to determine which channels are occupied. (2) Querying the modem status on any modem cards in the chassis to determine which modems are occupied. (3) Querying the LED status and gauging utilization from the chassis LEDs being displayed on the cards. (1) DS0 status on line cards This would be completely accurate (providing you queried all line cards). However, you do need to adjust the object you use by the line card type: Per-Channel Bulk --------------------------------------------------- Dual T1(*) ds0StatDs0 ds0BulkAccessStatDs0Modem Dual PRI ids0StatDs0 ids0BulkAccessStatDs0Mdm HDM usrds0StatDs0 usrds0BlkAccessStatDs0Mdm(+) (*) This is either the original 186 dual T1 card or the newer 386 dual T1/PRI card running the T1 code. (+) This object is not yet populated properly on the HDM code base. It's also technically in the DS1 mib (rds1.mib) rather than the DS0 mib even though the prefix is wrong. You'll find that the per-channel objects are very slow to query on the Dual card (slow, but not unbearably so on the HDM) so I'd strongly suggest the bulk objects, which basically have the normal status objects for every channel encoded byte by byte. (2) Querying the modem status You can use the "mdmCsStatus" object for each modem in the chassis (quad and HDM) to get the current status and judge those that aren't currently idle as in use. The actual response by the modems is quite fast (unlike the per-channel DS0), but there are a lot of these objects to query and no "bulk" equivalent. However, if you cluster a large number of modems into a single SNMP PDU you can probably get status for a doubled up chassis (48 modems in quads, and another 48 in two HDM cards) in under 15 seconds. (3) Checking LED status This is actually the fastest method, but with the introduction of HDMs it loses on exactness. Using the chassis LED objects: uchasFrontPanelLedColor, uchasFrontPanelLedColor2 uchasFrontPanelLedStates, uchasFrontPanelLedStates2 you can query all of the LEDs in the chassis. If you know which cards are modems, you can use an amber LED for training and a green for online to count users. (The *2 objects are for cards with more than 12 LEDs, like the HDM). The breakdown of these objects are documented in the NMC MIB reference - or at least for the main objects, since the *2 objects are newer, but they follow the same format just for LEDs > 12. The rub is with the introduction of HDMs. Since they don't have an LED for each modem (but rather one per 10% of the modems) you have to guess a bit with HDMs in a chassis. But you can't beat the speed, since you're just querying 4 objects and can do so in a single PDU. So what do I do? For my purposes, I have a "show chassis" command in one of our tools that presents a front view of a chassis (cards, LEDs, etc..) and it uses the LED information to quickly summarize usage as in my point (3). For HDMs, I'm forced to estimate somewhat, although prior to that point it was an exact count. The other more typical approach we use is from (1) for the DS0 status, since it lets us get a summary view of all input channels into a hub. For that, I've done code that always tries the bulk query (selecting the appropriate one for the card(s) in question) first, and then issues appropriate channel queries for any channels not properly returned in the bulk object. I've found that just checking for 0 values in the bulk object serves well here, since it covers both cases where the NMC supports the bulk object but the card does (earlier code) as well as issues where the card doesn't support it properly (like current HDMs). Oh, and some versions of the dual E1 PRI code incorrectly leaves out the 32nd channel on a span in the bulk object, so this picks up that and requeries it directly. Unfortunately, this isn't like a single object to query to give this summary. That's mostly because the TC is a distributed architecture, where the line, modems and terminal server functionality is distinct, so asking a question like "how many users are on" is more ambiguous than you might think, since the answer may differ depending on which card you are talking to and what it considers idle or "on". In some cases, what you probably want more than anything else is the NETServer view of online users, but unfortunately it doesn't do SNMP anywhere near as well as the rest of the chassis in terms of providing access to all of its information. So to work with MRTG you'd need to encapsulate this intelligence in a script or program that could then output the right value. My guess is that the LED poll would probably be suitable for your purposes (particularly if you don't have HDMs yet). Oh, and I just realized that there was a similar request last year that I also answered, so I'll include that response below (although it doesn't cover HDMs since they weren't out at the time) for added reference. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/ - - - - - - - - - - - - - - - - - - - - - - - - - In-Reply-To: Your message of Wed, 30 Jul 1997 16:16:49 -0400 Message-ID: <CMM.0.90.2.870336536.db3l@foo.ans.net> Bosco Tsang <bosco@ipoline.com> writes: > Is it possible for me to get a user count for USR Total Control via SNMP, > so that I can plot it out via MRTG? What are the OIDs? I'm not sure how implementable these methods are with MRTG (is that an SNMP display tool of some sort?), but if all you're really concerned with is simultaneous port count (as opposed to username information or other NETServer authentication type of stuff), there are a few approaches you can take. The NETServer itself won't help all that much (it supports very little via SNMP) but you can try: * Walking the mdmCsStatus object. This will give you a list of objects for each modem (e.g., mdmCsStatus.2001 is for modem 1 in slot 2). Count those that have an enumerated value of onlineAnswer(8) - or onlineOriginate(7) too if you do outbound, and you'll have a count of "live" modems. You can look at other enumeration values to see users that are training, or in other states. It's not the fastest in the world, but walking a hub with 48 modems should take about 10 seconds or so. You can "cheat" if you know just how many modems there are, by constructing a query that uses GetNext, and queries the objects just in front of the known modems. For example, if you know you have 48 modems in slots 2-13, instead of doing 48 GetNext queries to walk table dynamically, build a single query that queries the objects: mdmCsStatus.2000, mdmCsStatus.2001, mdmCsStatus.2002, mdmCsStatus.2003, mdmCsStatus.3000, mdmCsStatus.3001, (etc...) mdmCsStatus.13002, mdmCsStatus.13003 and then issue a single GetNext query (or just a few to avoid making the PDU too large). The instances will "roll" to the appropriate modem (e.g., your .2000 query will get you modem 2001), and you'll have many less PDUs to handle, cutting down the overhead. You can query a hub worth of modems this way in only about 2-3 seconds. Note that this approach is only useful for hubs where the modems handle all callers, so if you have ISDN hubs that pass callers right to the NETServer this won't count them. * Ask your circuit card(s) what the status of their individual span channels are. You can break it out by those that are connected in or out, busied out, etc.. any of the enumerations as shown in the MIBs for ds0StatDs0 (channelized T1) or ids0StatDs0 (PRI). One approach is to actually walk the appropriate MIB tree but I wouldn't recommend it if you can parse an opaque object instead, because walking the DS0 status tree is deathly slow, even trying the trick above - the interface to the T1/PRI cards is just too slow. Instead, I would suggest querying the newer bulk access objects (ds0BulkAccessStatDs0Modem for channelized configurations and ids0BulkAccessStatDs0Mdm for PRI). These objects are a single opaque value with information on all channels in one query, and allow you to query the span information in a single SNMP object in well under a second. The instance for the objects is the entity of the DS1 span (e.g., 1001 for the first DS1 in slot 1). You do have to break the information apart a bit, once you get it back, but for the performance it's worth it. For the channelized T1 case the object is two bytes per channel. The first byte is the ds0StatDs0 enumeration and the second the ds0StatModem enumeration. So just pluck out every other byte for your DS0 channel status. In the PRI case, the bulk object is 4 bytes per channel, the first being the ids0StatDs0 enumeration, the second the ids0StatDevConnTo enumeration, followed by ids0StatSlotConnTo and ids0StatChanConnTo (the latter two are actually 0xFF for no connection which isn't really documented in the objects themselves). So pluck out the first of every 4 bytes and you've got your DS0 status. Once you've got the status of your channels you can break them out, and sum them up however you'd like. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 789-5327 | / 100 Clearbrook Road, Elmsford, NY 10523 \ Fax: (914) 789-5310 \ \-----------------------------------------------------------------------/ - - - - - - - - - - - - - - - - - - - - - - - - - In-Reply-To: Your message of Tue, 9 Dec 1997 21:19:44 -0800 (PST) Message-ID: <CMM.0.90.2.881738093.db3l@valheru.ny.ans.net> Jaye Mathisen <mrcpu@cdsnet.net> writes: > Heck, I haven't even figured out who the stupid things are named. Or > how to take CHS.MIB and extract the full name for the > uchasFrontPanelLedStates data. The LED objects reside in the uchasConfig tree, with a full path from the root of: iso.org.dod.internet.private.enterprises.usr.nas.chs.uchasConfig The full numeric OIDs for each of the objects are: uchasFrontPanelLedStates (1.3.6.1.4.1.429.1.1.3.5) uchasFrontPanelLedColor (1.3.6.1.4.1.429.1.1.3.6) uchasFrontPanelLedStates2 (1.3.6.1.4.1.429.1.1.3.8) uchasFrontPanelLedColor2 (1.3.6.1.4.1.429.1.1.3.9) The *2 objects are for the NMC 5.x code and support cards that have more than 12 LEDs (like the HiPer DSP). These objects are just opaque octet strings, which you have to decode yourself. You can find a description of the breakdown in the NMC 3.0 SNMP MIB documentation - I've included an early draft that I happen to have online. About the only correction is that I've found that the color object does in fact indicate the LED color for both power supplies as the first two LEDs in the last bank of octets. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/ - - - - - - - - - - - - - - - - - - - - - - - - - 4.3. THE CONFIG GROUP Most of the MIB objects in this group are self explanatory. The exceptions would be uchasFrontPanelLedStates, uchasFrontPanelLedColor and uchasNicStates. In general, these objects are bit maps. The current design handles up to 12 LEDs per card. Each LED's information is encoded in a nibble (4 bits), therefore each card takes up 6 bytes out of the octet string and 17 slots requires 102 bytes in the octet string. It is the responsibility of the management station to know how many LEDs each kind of card has on its front panel. The MIB does not identify how many of the 12 nibbles for each card carry useful information. Power supply LEDs are not included in this object. Their LED status is to be deduced by the PSU state (see power supply section below). In addition to LED state and color, a few of the bits in these objects were used for additional information. The bit definition for these three objects is shown below. uchasFrontPanelLedStates Each LED uses a nibble to indicate its current state. There are 12 nibbles (6 bytes) per card. The most significant nibble of the first byte is the first LED on the first card. The most significant nibble of the seventh byte is the first LED on the second card and so on. There is no indication in the MIB how many nibbles are valid for a given card. Let's assume we number the bits in a nibble as b3=MSB...b0=LSB. Of each nibble, the three least significant bits are defined as follows... b2 b1 b0
Subject: Re: (usr-tc) Bug Reports and tech support
From: Jay Nitikman <jay@cruzio.com>
Date: 1998-02-24 19:48:46
My only real complaint about USR tech support is that when I have a problem with my T/C they ask me to drop my firewall so that they can log into the hub and fix it. They have solved my problems this way but I have not learned how they fixed it, so the next time a similar problem occurs I have to call them backup. I'd prefer that they talk me thru the procedure, kinda like the control tower talking passenger into landing a crippled plane. They seem to be in a hurry to move to the next call. -- Jay Nitikman (jay@cruzio.com) Cruzio is a mom and pop Internet Service Provider Web: http://www.cruzio.com Email: info@cruzio.com Voice: 423-1162
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 19:52:15
barberw@tidewater.net writes: > ifNumber.0 is a count of current interfaces on the Netserver and one > needs to be subtracted because the LAN port gets counted. The only thing to recognize by querying this is that it only counts the number of active IP interfaces (or SLIP/PPP users that have properly authenticated and/or negotiated a session). It won't count users currently signing on (although they do tie up a port), someone sitting at the login: prompt, or other users (telnet/rlogin/netdata). You can pick up the telnet/rlogin/netdata users by walking the TCP connection table though (also part of MIB-II) but that won't pick up users that haven't actually made their connection yet. Of course, if your service is just a SLIP/PPP service and the issue of busy ports prior to session negotation isn't too big (probably a reasonable assumption) then this can work fine, and it's definitely a simple, single query. Oh, since it wasn't mentioned - this query is to the NETServer directly, and not via the NMC (which all the previous SNMP queries have been to). So you do need to ensure that you've activated SNMP on your NETServer and configured it for appropriate access, independent of the NMC configuration. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) restricting connect rates...
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 19:58:18
Jay Nitikman <jay@cruzio.com> writes: > I'm sorry if this has been covered before but I cannot find the answer > in the archives. I want to restrict modem connections to X2 only. There should be a note from me in the archives on this, but I'll include the relevant table below. Unfortunately I think in terms of actual MIB objects, not the naming that TCM seems to use, but I'll try to indicate what seems to be appropriate mappings - maybe the TCM stuff also includes the MIB names somewhere for a double check. Controlling the objects this way will cause any users calling in from something other than an x2 modem to fail to train: - - - - - - - - - - - - - - - - - - - - - - - - - All you need to do is program the modems involved not to work with modulations other than x2. You can do this by changing some of the variables in the modem signal control table (which should be accessible by that name in TCM, or in the mdmScTable table if you are using the MIB directly). The objects you want to change are: Object(s) Set to ---------------------------------------------------------------- mdmSc300, mdmSc1200, mdmSc2400 enable Disables lowest baud rates. The value is actually enabled since the objects are defined as a negative (if enabled they prevent the specified connection rate). mdmScHstMod disable Disables USR proprietary HST mode. mdmScV32Mod, mdmScV32Bis, mdmScV32BisEnhance, disable mdmScV32TerboModeEnable Disables V.32 and variations (9600-19.2K) mdmScVFCModeEnable disable Disables V.FC mdmScV34ModeEnable disable Disables V.34 However, you should ensure that the mdmScV34pModeEnable variable is enabled (the default), since it comes into play when computing the back-channel for x2 calls - but having it enabled does not allow V.34+ connections since that requires the V.34 mode object to be enabled. - - - - - - - - - - - - - - - - - - - - - - - - - I think that these correspond to the following TCM objects: > 300 Baud (S48.0) disable > 1200 Baud (S48.1) disable > 2400 Baud (S48.2) disable Enable these. > HST Modulation (S13.5) enable Disable this. > V.32 Modulation (S27.2) enable > V.32 bis Modulation (S34.0) enable > V.32 Enhanced Mode (S34.1) enable > V.32 Terbo Modulation (S34.7) enable > V.34 Modulation (S56.6) enable > V.FC Modulation (S56.7) enable Disable these. And you should be set - let everything else remain as defaulted after initialization the modem to defaults. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) restricting connect rates...
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 20:01:38
Pete Ashdown <pashdown@xmission.com> writes: > >300 Baud (S48.0) disable > >1200 Baud (S48.1) disable > >2400 Baud (S48.2) disable > >High Speed (S48.3) disable > > disable all Careful - these objects are kind of opposite to what the intuitive meaning is (to me at least). They are defined as "restriction" objects which means that you need to enable them to actually disable the modulation. But, high speed is a global control, so you do want to leave it disabled (because enabling it, or restricting high speed connections, would also prevent x2). -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) restricting connect rates...
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 20:03:22
Brian <signal@shreve.net> writes: > > >V.8 Mode (S54.7) enable > > >V.8 Call Indicator (S54.6) disable > > you sure v.8 isn't needed? (I think you zapped the part where the previous message said to disable all, since as shown above V.8 is enabled, which is fine. The call indicator isn't necessary) You definitely need V.8 for V.34 or higher speed connections. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) restricting connect rates...
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 20:09:15
rpanula@dacmail.net (Russ Panula) writes: > >Minimum High-speed Direction Link Speed (&U) variable > > Set this to the lowest X2 speed (33.333bps). > > That alone should limit the connections to X2... Well, technically x2 can go lower than 33.3K (down to almost 25K, although yeah, the odds of ever getting anything lower than 33.3K is darn low), but even more importantly, V.34 can get you 33.6K on good lines, so you'd need to raise this to at least 37.3K to be sure of excluding all non-x2 users, but then you might exclude some of the poorer quality x2 users. Not to mention this will cut off all users who do have x2 modems, but who have to fall out to V.34 rates (either during the initial training or at some point during the call). The former might not be so bad, if you don't want them using the line anyway, but the latter would result in a disconnect rather than a retrain down to V.34 - which the user would perceive as a good connection with a spontaneous disconnect. I've also seen problems with various client code versions when they try to implement rate limitations transmitted by the server, so I'm not a big fan of production use of this sort of method - rather considering it more of a client side workaround when trying to deal with an overly aggressive x2 training condition. Controlling access at the modulation level cleanly cuts off all users who are not using a modem that is capable of x2 - even if in the end they train or retrain out of x2. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: (usr-tc) [Q] USR Total Security Server using 100% of CPU as Service
From: DWX Network Operations <tracy@dwx.com>
Date: 1998-02-24 20:30:12
Greetings. When running USR Total Security and Accounting server as a service under NT 4.0, I find that my CPU runs right at 100% when I check the Task Manager. Under the list of processes I see that radserver.exe fluctuates between 90 and 95 percent of the CPU usage. When I kill the service and run the server as an application my CPU usage is about what I would expect (8 to 10 percent). Is this a problem for anyone - or a documented scenario that I've overlooked? I, of course, would prefer to run this as a service - any words of wisdom? Thanks, Tracy Hinshaw, DWX NOC noc@dwx.com
Subject: Re: (usr-tc) Sample quad configuration..
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-24 20:32:01
On Tue, 24 Feb 1998, Pete Ashdown wrote: > System Administrator said once upon a time: > > > >Can someone post a sample quad configuration that will accept ISDN calls? > >I tried setting my ISDN call control options to the *exact* same thing as > >my HDM ISDN call control, and now ISDN can't connect to the quads (yes, > >I've got ISDN-GW set to 0). I'm trying to get better performance from the > >quads than I do from the Munich, and not having much luck. > > What version are your Quads on, and what are those settings specifically? Currently, 5.6.7 (have the upgrade, haven't done 'em yet). I *did* get ISDN to connect, but I'm having a horrid problem (had this problem before as well). For some reason, when I put ISDN calls on to the quads, latency in general goes through the roof -- but ONLY for ISDN connections. Not talking about UDP latency here, either. For example, my average ping time across ISDN on quads is 100ms, instead of 50 or so when I'm on the Munich. Telneting becomes painful and "spurty", whereas it's perfectly smooth on the Munich. Exact same thing happens to ISDN calls on HDMs. I keep hearing on this list how it's tres important to get your ISDN on to quads, in terms of improving Quake/UDP latency issues, and I would *LOVE* to do so if it weren't for this horrid overall latency problem. As a footnote... my solution to ISDN on the quads was to *ENABLE* Async/Sync PPP conversion. However, enabling this on HDMs seems to have the opposite effect (ISDN calls don't work). Very odd. For reference my Quad ISDN Call Control config is: V110 Rate Adaption : disable Force Fixed Network Rate : notForced Force Network Rate Speed : kbps56 Enable 45-65 Second Link Delay : noDelay Analog Calls Over Digital : enable Async PPP/Sync PPP Conversion : enable X.75 : disable Set Date Mode of Modem : autodetect Set Originate HDLC Protocol : none Set Originate non-HDLC Protocol: none Set Originate Analog Modem/Fax : analogModemFax V120 : disable X75 Frame Size : 2048 X75 Window Size : 2 Regards, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Radius and ptpp
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-24 21:20:12
On Wed, 25 Feb 1998, Richard Bosire wrote: > Hi Guys , > > Does anyone know which version of Radius ptpp and what version of the > netserver code this will > with .. > > TIA > > cheers > > bosire Every Radius will support PPTP, see in the dictionary file under the framed-protocol section you will see something like this VALUE Framed-Protocol PPP 1 now go ahead and add a new item there VALUE Framed-Protocol PPTP 3 and when you add users you should say Framed-Protocol is PPTP krish > ] > -- > > \\|// - ? > (o o) > +==================================oOOo=(_)=oOOo========+ > | Richard Bosire rbosire@africaonline.co.ke | > | AfricaOnline Ltd | > | union towers, 2nd floor | > | tel: 254-2-243775 | > | .oooO | > | http://www.africaonline.co.ke ( ) Oooo. | > +===================================\ (==( )==========+ > \_) ) / > (_/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Quakeworld and Quake2 lag
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-24 22:32:49
I've never heard this. I have a pool of exactly 48 ip addresses for each of my three racks, and my parent company has done the same thing in another couple dozen racks. We have never needed even one more IP address in the pool. Steve -----Original Message----- >BTW, USR told me that you need to set the IP pools to 15% more IP >addresses than the system needs to service the modems in your hub. >Something to do with the time needed to release the IP addresses back >into the pool upon disconnection. > >Mike wrote: >> >> After you set up the pools and the limit did you save all & reboot? The ip >> pool and limit only take effect after >> a reboot.. This has caused many people some grief.. >> >-- >======================================================= >=========== Andrew Aken - President ========= >====== GlobalEyes Communications, Inc. ====== >=Southern Illinois' Fastest Connection to the Internet= >========== http://www.GlobalEyes.net ======== >======================================================= > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Bug Reports and tech support
From: David Bolen <db3l@ans.net>
Date: 1998-02-24 23:02:20
Jay Nitikman <jay@cruzio.com> writes: > I'd prefer that they talk me thru the procedure, kinda like the control > tower talking passenger into landing a crippled plane. Just cheat... give them an address that isn't the TC itself but some forwarding daemon that just logs the information and transparently forwards it on to the appropriate rack. :-) Or sniff (tcpdump or whatever) the ethernet so you can see what they are sending to the TC - although that might be a pain to parse. Since you've already got a firewall around, you may find that you can set up the firewall to log the information that it is forwarding, which you can look into later. It's easiest to follow TCP text based stuff like the NETServer logins, but you can forward SNMP just about as easily and decode the packets along the way to see what variables are being requested or set and with what values. It's a passive operation, so the techs can do whatever they want and you just get a log of it to peruse later. We use stuff like this for all sorts of types of gear we use, including such items as CSU/DSUs on the backbone. Since we won't open up network filters in general, but sometimes you just need to let the vendor take a peek, we provide forwarders for vendors from machines that they can connect to which in turn connect to the real units. In some cases we do it with a dialup (the vendor calls a modem they think is connected right to their gear, but it's really connected to a listening daemon that then forwards the traffic over a secured TCP connection to the real device) and in others just with a daemon on an exposed host, but in each case we can log whatever happens through the link, and it is transparent to the vendor. Heck, I even did stuff like this when I was managing our local site PBX, and the vendor came in to do some work - I had everything logging to a separate file that I could look at later to learn from what they did without slowing them down at the time. I agree it would probably be nicer to be walked through step by step, but if you can't have that, you can always just grab the information :-) -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Bug Reports and tech support
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-24 23:12:03
Thus spake Jay Nitikman >My only real complaint about USR tech support is that when I have a >problem with my T/C they ask me to drop my firewall so that they can log >into the hub and fix it. They have solved my problems this way but I >have not learned how they fixed it, so the next time a similar problem >occurs I have to call them backup. Amen, amen, and amen. I'm about to start refusing to give the techs any access and just tell them that if they can't help me solve it then they need to escalate the call to someone who can. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) Sample quad configuration..
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-24 23:13:23
>> System Administrator said once upon a time: >> > >> >Can someone post a sample quad configuration that will accept ISDN calls? I have denoted the differences in your config and ours with "<--" Good Luck.. >V110 Rate Adaption : enable <-- >Force Fixed Network Rate : notForced >Force Network Rate Speed : kbps56 >Enable 45-65 Second Link Delay : noDelay >Analog Calls Over Digital : enable >Async PPP/Sync PPP Conversion : enable >X.75 : enable <-- >Set Date Mode of Modem : autodetect >Set Originate HDLC Protocol : none >Set Originate non-HDLC Protocol: none >Set Originate Analog Modem/Fax : analogModemFax >V120 : enable <-- >X75 Frame Size : 2048 >X75 Window Size : 2 > am _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) Sample quad configuration..
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-25 00:13:39
At 08:32 PM 2/24/98 -0500, System Administrator wrote: >As a footnote... my solution to ISDN on the quads was to *ENABLE* >Async/Sync PPP conversion. However, enabling this on HDMs seems to have >the opposite effect (ISDN calls don't work). Very odd. > We have it enabled on both quads and hdms.. I thought it was the hdms you were having isdn troubles with and not the quads.. And enabling it on hdm's was the fix.. maybe i have it bassackwards.. Anyway should work on both *enabled*.. am _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) mdmCsStatus
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-25 00:50:29
At 07:42 PM 2/24/98 EST, David Bolen wrote: [snip] >FYI, for those who may be as crazy as me, I optimize this estimate for >HDM cards slightly, because of the behavior that the first LED >actually turns on with the very first modem (and not after 2-3) and >the last LED only turns on when _all_ the modems are in use. So while >I treat most active LEDs as 10% of the total available modems >(according to card and circuit type), the uppermost active LED (as >long as it isn't one of the top 2 LEDs of the card) is considered to >only count half as much of the standard per-LED value, or 5% of >total. > >This is to allow for the fact that the uppermost LED turned on at the >bottom of the 10% threshold but might not include the full set of >channels. The uppermost two LEDs are excluded since the top LED only >comes on when full, and because of that the "gap" from the second to >top LED is pretty much two full LED values, so when the second to top >LED is on, guessing at half of that two full steps (which is really >just a single normal step) is a good estimate. Get that? :-) > [snip] Pardon me, but as a person new to all this, I find this all sort of comical.. At least it would be funny if pmwho/pmmon worked with arc/hdms and I *knew* what my usage was. I'm currently waiting on brian to put something together. It just seems that a better way of looking at whos on line and how many modems are in use from snmp.. I mean what are the first two things admins need??.. Calculating usage off of led's is crazy, but I love the elegant understanding David has given the issue. I guess if you are going to kludge something, might as well do it right! I guess I don't consider this kludge on quads as much as hdms however david's hdm calculations seem accurate enough for what I need.. But telneting using expect's, snmping led's, and all, might have been ok with older generation quads/netservers, but in building the Hiper stuff, they had a chance to make *all* necessary improvements. Why not a cleaner snmp interface for the *most common* stuff??.. Maybe the servers don't really know how loaded they are (stored in a register easy to query) and they didn't want to track that for no reason other than quick snmp queries.. Maybe what I want is a HiperNMC! Sure would be cool to set a trap for 95% usage on the last hub or something.. And pull usage in a single query that even I could handle.. MRTG graph of usage would be nice also. Sorry for rambling but I don't feel like watching tcm tonight... (reruns) am _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: Re: (usr-tc) mdmCsStatus
From: David Bolen <db3l@ans.net>
Date: 1998-02-25 02:27:41
Allen Marsalis <am@shreve.net> writes: > Pardon me, but as a person new to all this, I find this all sort of > comical.. Well, I did say you'd have to be as crazy (insert favorite adjective here) as me :-) However, on a slightly more serious note, I hope the recent spurt of discussions on this hasn't obscured the fact that there are in fact more formal ways of getting accurate information, based on your requirements and constraints. There may not be a single solitary object that holds the aggregated data you want, but I would argue that there are lots of ways to aggregate this data, and different people might have different requirements. For 3Com to include all possible aggregations in their MIB is infeasible, and I'd rather that they included access to the most granular data and permitted me to do my own aggregation than not. It could still be useful to add "typical" aggregations, but not to my mind a requirement. A lot of this arises because in many ways the Total Control hub is a container for a variety of independently operating units, which coordinate their activity but in many ways are still independent. That has advantages (modularity, etc..) but disadvantages sometimes when trying to get a bird's eye view of what is going on. > It just seems that a better way of looking at whos on line and > how many modems are in use from snmp.. I mean what are the first > two things admins need??.. The mdmCsStatus table will definitively tell you this, so if your question is "what modems are in use" that's the right answer. Of course, as in many things, it's not necessarily the only way to get at the general question that "what modems are in use" might stand in front of which might be a more general utilization question. For what it's worth, walking 20 objects at a time you can query all the modem status for a doubled-up chassis (96 modems) in about 42 seconds. Not exactly instantaneous but not years either. Set up that query as a script for MRTG and you can monitor your utilization on a reasonably real-time level. If instead you poll the DS0 channel status using bulk objects for the Dual-T1/PRI card, and the normal objects for the HDM cards, you can determine status of all of your channel in 11 seconds today (which should go down to about a second or two once the HDM code supports the bulk query). > Calculating usage off of led's is crazy, > but I love the elegant understanding David has given the issue. I > guess if you are going to kludge something, might as well do it > right! I guess I don't consider this kludge on quads as much as hdms > however david's hdm calculations seem accurate enough for what > I need.. (For what it's worth, you can make an exception for no LEDs and all LEDs for the HDMs in which case you're perfectly accurate - no modems and all modems respectively) Also note that the LED discussion is but one approach to the issue, which just happens to come up because there is a convenient and fast object for the LEDs which can be munged to help provide a count. It's by no means the only method, nor even the most preferred depending on your need. For what it's worth, my LED computation came about just because I had a command to display a chassis LED status, and I happened to count modems while doing that for a quick summary display. I needed to support HDMs somehow in that, so I use my method (and display the total count with a tilde (~) for an estimate) in such a case, but it's not so much something that we would use for real time monitoring, which is based on our general statistics daemon which always knows which modems are in use. > But telneting using expect's, snmping led's, and all, might have > been ok with older generation quads/netservers, but in building the > Hiper stuff, they had a chance to make *all* necessary improvements. > Why not a cleaner snmp interface for the *most common* stuff??.. Well, one point that hasn't been made is that the HiPer ARC does have much more elaborate SNMP support than the NETServer, including dumping a list of users and such. It's not perfect (I've definitely got issues), but it's a good start at the moment - far better than the NETServer. Part of the issue with HiPer DSP is that even if they made *all* necessary improvements (and I would argue that *all* is a subjective term) you'd still have to deal with the existing MIBs in order to handle existing racks. For the most part the HiPer DSP stayed compatible with the existing MIB structure which to be honest probably helps more than it hurts - I know it did in our case. What I do think would be a good thing for 3Com to take from this discussion (for those who may be reading) is that adding some additional MIB tables - which may be maintained by the NMC itself - to provide a few very common aggregation levels of various sorts of "hub level" information, would probably go over very well in situations when looking for straight forward hub-level SNMP objects that can be utilized without too much coding. After all, you can do any of the above queries even with MRTG, but it requires more work than a nice convenient object does. Of course, there's the rub - just exactly what should that object contain, and how should it be summarized. By user, by call type, etc.. It's potentially a can of worms, but I do bet that even a few selections would go over well. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) mdmCsStatus
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-25 04:04:36
At 02:27 AM 2/25/98 EST, David Bolen throws pearls of wisdom: [snip] >What I do think would be a good thing for 3Com to take from this >discussion (for those who may be reading) is that adding some >additional MIB tables - which may be maintained by the NMC itself - to >provide a few very common aggregation levels of various sorts of "hub >level" information, would probably go over very well in situations >when looking for straight forward hub-level SNMP objects that can be >utilized without too much coding. After all, you can do any of the >above queries even with MRTG, but it requires more work than a nice >convenient object does. Would be nice. >Of course, there's the rub - just exactly what should that object >contain, and how should it be summarized. By user, by call type, >etc.. It's potentially a can of worms, but I do bet that even a few >selections would go over well. > Yeah especially the two or three items that *everyone* needs to monitor line count and users.. Sure, if we had it, some folks would complain about some missing trival bit. But for something as simple and straight forward as modem usage, what could the debate be about? Optimist: cs + 'modems_used' followed by 'total_modems' Pesimist: cs + 'total_modems' followed by 'modems_used' Government: cs + 64bit encryption key followed by 'modems_used' Bill Gates: cs + 'modems_used' behind new proprietary format OIDs BellSouth: cs + 'total_modems' AGIS: cs + 'total_spammers' internic: cs + First Virtual ID followed by 'modems_used' iamerica: cs + random_number_seed followed by 'modems_used' lucent: cs + license_number followd by 'modems_used' Sorry for the joke david but I just couldn't resist. In contrast to your eloquent and useful writing, I sometimes have to poke fun in order to be useful.. Thanks for your ideas and information which I'm sure will help brian come up with something that works well.. And I plan on taking a look at the aforementioned MIB reference myself "just for grins"... am _____________________________________________________________ Allen Marsalis President Voice: 318.222.2NET (2638) Shrevenet, Inc. mailto:am@shreve.net 333 Texas St. Suite 619 FAX: 318.221.6612 Shreveport, LA 71101 http://www.shreve.net _____________________________________________________________ Thoughtful Provider of Internet Services
Subject: (usr-tc) new v.90 modem problem
From: Richard Gamberg <bbhi@prodigy.net>
Date: 1998-02-25 07:53:45
There is at least *ONE* new USR v.90 modem in the field.... And the owner can no longer get x2 speeds on local call. ISP is Prodigy Internet. Access is on Split Rock POP. Who is Split Rock? Anybody know if they aren't using the latest quad-modem firmware? Or is this a show-stopping new bug? See http://pages.prodigy.net/bbhi/x2-4.htm Aloha, Richard (btw, the # the user is calling is listed by Prodigy as 56k at http://www.prodigy.com/learn/lrnindex.htm MINNEAPOLIS MN 612 252-2170 56K X2 (All my 56k ramblings start at: http://pages.prodigy.net/bbhi/r-rnut-x2.htm )
Subject: (usr-tc) Radius and ptpp
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-25 08:06:02
Hi Guys , Does anyone know which version of Radius ptpp and what version of the netserver code this will with .. TIA cheers bosire ] -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: (usr-tc) new v.90 modem problem
From: Richard Gamberg <bbhi@prodigy.net>
Date: 1998-02-25 08:06:50
There is at least *ONE* new USR v.90 modem in the field.... And the owner can no longer get x2 speeds on local call. ISP is Prodigy Internet. Access is on Split Rock POP. Who is Split Rock? Anybody know if they aren't using the latest quad-modem firmware? Or is this a show-stopping new bug? See http://pages.prodigy.net/bbhi/x2-4.htm Aloha, Richard (btw, the # the user is calling is listed by Prodigy as 56k at http://www.prodigy.com/learn/lrnindex.htm MINNEAPOLIS MN 612 252-2170 56K X2 (All my 56k ramblings start at: http://pages.prodigy.net/bbhi/r-rnut-x2.htm )
Subject: Re: (usr-tc) Radius Attributes...
From: MegaZone <megazone@megazone.org>
Date: 1998-02-25 08:18:52
Once upon a time Scott Portmaster List shaped the electrons to say... >I Don't happen to have the USR Tc books with me at the present time (At >home), does anyone remember what attribute 25,32 are? Just check the RFCs - the Lucent RABU web/ftp site is also the IETF archive site. -MZ -- <URL:mailto:megazone@megazone.org> Gweep, author, webmaster, human being, me "A little nonsense now and then, is relished by the wisest men" 508-791-9803 <URL:mailto:megazone@gweep.net> <URL:http://www.megazone.org/> Hail Discordia!
Subject: Re: (usr-tc) Netserver idle-timeout
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-25 08:19:34
Pete Ashdown wrote: > Richard Bosire said once upon a time: > > >Hi Pete ,No , I have been trying it from the netserver command prompt .. > >Pliz you dont mind shedding some ideas on how to do this from RADIUS > > I just have the following line in my DEFAULT user entry: > > Idle-Timeout = 600 > Hi ..I dont have any entry for all the ppp clients .. Does this imply i have include all my ppp clients in the /etc/raddb/users file?.. cheers > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: Re: (usr-tc) Netserver idle-timeout
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-25 08:41:50
Thanx ,,,,.. I'll give it a shot and see it behaves .. bosire Jason_Kelton@3com.com wrote: > Richard, > > You might want to try upgrading to NETServer 3.7.24... and while you're at > it, you might need to upgrade your NETS with 16 Meg of RAM. > > Regards, > > Jason. > > bosire@ns1.africaonline.co.ke on 23/02/98 19:51:23 > > Please respond to usr-tc@lists.xmission.com > > To: usr-tc@lists.xmission.com > cc: (bcc: Jason Kelton/AU/3Com) > Subject: (usr-tc) Netserver idle-timeout > > Hi .. > I have this problem of netserver idle-timeout ,, The suppossedly > working feature does not work on my setup .. > Is anyone having the same problem , is there any work-around .. > my netserver details are: > U.S. Robotics > Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.3.28 > Build date: Dec 13 1996 > Build time: 13:54:59 > Network Interface Card: Ethernet & Frame Relay Combination (26) > ISDN Interface Card : MUNICH32 (4) > Packet Bus Circuit : Enhanced > Licensed for 60 ports. > ] > cheers and thanx > -- > \\|// - ? > (o o) > +==================================oOOo=(_)=oOOo========+ > | Richard Bosire rbosire@africaonline.co.ke | > | AfricaOnline Ltd | > | union towers, 2nd floor | > | tel: 254-2-243775 | > | .oooO | > | http://www.africaonline.co.ke ( ) Oooo. | > +===================================\ (==( )==========+ > \_) ) / > (_/ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: (usr-tc) Q: Performance on Chan T1
From: Phil Freed <phil@freed.com>
Date: 1998-02-25 08:46:27
I posted this a couple of weeks ago, but I got no responses. I thought it worth one more shot. Or - are there any recommendations on a better forum for this question? This much is old hat, I suspect: We have found that users with 33.6 modems cannot achieve 33.6 when they connect through a channelized T1 using AMI (as opposed to B8ZS). We've been told that 33.6 connections are impossible under these conditions because of the bandwidth lost to Robbed Bit Signalling. But here's where it gets more interesting: But we've also been told by folks with PM3s (well, at least one major ISP) that they regularly get 33.6 connections under these circumstances. Of course, connection speed isn't the most important thing, since it can always be renegotiated. Still, it _is_ what most of our users see, and it's what we get lots of support calls on. Given all of this, I was wondering what experience others have using various kinds of lines - especially channelized T1s - on TCs or other NASs. I'd be interested in hearing about both connection speeds and negotiated speeds. In fact, if folks just want to dump stats, then send them directly to me: pfreed@cybertours.com and I'll summarize for the list in a week or so. While I'm on the subject: I've been told by a USR engineer that you will seldom (if ever) see the upstream portion of an X2 connection go over 28.8. Is this true? Anyone have any thoughts on why it might be so? Thanks for your time. --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."
Subject: Re: (usr-tc) Netserver idle-timeout
From: Richard Bosire <bosire@nairobi.africaonline.co.ke>
Date: 1998-02-25 08:47:13
Hi Kelton .. At last someone from 3com( or is it usr) has come to my rescue .. Thanx ... While we are still on this issue , have a few questions to ask .. Is it possible to disable someone from making more that one simulatenous ppp connection?. What version of Radius supports ptpp and what version of netserver code does this work with.. cheers and thanx bosire Jason_Kelton@3com.com wrote: > Richard, > > You might want to try upgrading to NETServer 3.7.24... and while you're at > it, you might need to upgrade your NETS with 16 Meg of RAM. > > Regards, > > Jason. > > bosire@ns1.africaonline.co.ke on 23/02/98 19:51:23 > > Please respond to usr-tc@lists.xmission.com > > To: usr-tc@lists.xmission.com > cc: (bcc: Jason Kelton/AU/3Com) > Subject: (usr-tc) Netserver idle-timeout > > Hi .. > I have this problem of netserver idle-timeout ,, The suppossedly > working feature does not work on my setup .. > Is anyone having the same problem , is there any work-around .. > my netserver details are: > U.S. Robotics > Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.3.28 > Build date: Dec 13 1996 > Build time: 13:54:59 > Network Interface Card: Ethernet & Frame Relay Combination (26) > ISDN Interface Card : MUNICH32 (4) > Packet Bus Circuit : Enhanced > Licensed for 60 ports. > ] > cheers and thanx > -- > \\|// - ? > (o o) > +==================================oOOo=(_)=oOOo========+ > | Richard Bosire rbosire@africaonline.co.ke | > | AfricaOnline Ltd | > | union towers, 2nd floor | > | tel: 254-2-243775 | > | .oooO | > | http://www.africaonline.co.ke ( ) Oooo. | > +===================================\ (==( )==========+ > \_) ) / > (_/ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. -- \\|// - ? (o o) +==================================oOOo=(_)=oOOo========+ | Richard Bosire rbosire@africaonline.co.ke | | AfricaOnline Ltd | | union towers, 2nd floor | | tel: 254-2-243775 | | .oooO | | http://www.africaonline.co.ke ( ) Oooo. | +===================================\ (==( )==========+ \_) ) / (_/
Subject: (usr-tc) (USR-TC) [Q] USR TOTAL SE
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-25 08:50:00
I had the same problem until I stopped running it as a service. I run it as a task in my statup folder. No problems since. Jeff Binkley ASA Network Computing U>Greetings. U>When running USR Total Security and Accounting server as a service U>under NT 4.0, I find that my CPU runs right at 100% when I check the U>Task Manager. Under the list of processes I see that radserver.exe U>fluctuates between 90 and 95 percent of the CPU usage. U>When I kill the service and run the server as an application my CPU U>usage is about what I would expect (8 to 10 percent). U>Is this a problem for anyone - or a documented scenario that I've U>overlooked? I, of course, would prefer to run this as a service - any U>words of wisdom? U>Thanks, U>Tracy Hinshaw, DWX NOC U>noc@dwx.com U>- U> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" U> with "unsubscribe usr-tc" in the body of the message. U> For information on digests or retrieving files and old messages send U> "help" to the same address. Do not use quotes in your message. CMPQwk 1.42 9999
Subject: (usr-tc) RE: (USR-TC) [Q] USR TOTA
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-25 08:50:00
Aside from slowing everything else down on the machine it isn't a problem. I got around the problem by running it as a task and not a service. I am getting ready to test version 5.0 now. I'll let you know what I find. Jeff Binkley ASA Network Computing U>On Tue, 24 Feb 1998, DWX Network Operations wrote: U>> Greetings. U>> When running USR Total Security and Accounting server as a service U>> under NT 4.0, I find that my CPU runs right at 100% when I check the U>> Task Manager. Under the list of processes I see that radserver.exe U>> fluctuates between 90 and 95 percent of the CPU usage. U>Yes this is actually a problem with NT. You can open a DOS window and U>get the CPU to 100 %. I know that 4.x version of Radius server did U>show up 100 %. You have two work arounds, U>1. From the task manager you can prioritize the radserve.exe process U>to be set as high - which helps a little. U>2. Start the process as an application. U>I am not sure but you can try 5.x radius server - since there was ER U>release of security code that did lower the CPU level. U>However, 100% cpu does not cause any problems - nor have I heard of U>any U>krish U>> When I kill the service and run the server as an application my CPU U>> usage is about what I would expect (8 to 10 percent). U>> Is this a problem for anyone - or a documented scenario that I've U>> overlooked? I, of course, would prefer to run this as a service - U>> any words of wisdom? U>> Thanks, U>> Tracy Hinshaw, DWX NOC U>> noc@dwx.com U>> - U>> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" U>> with "unsubscribe usr-tc" in the body of the message. U>> For information on digests or retrieving files and old messages U>> send "help" to the same address. Do not use quotes in your U>> message. U>- U> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" U> with "unsubscribe usr-tc" in the body of the message. U> For information on digests or retrieving files and old messages send U> "help" to the same address. Do not use quotes in your message. U> CMPQwk 1.42 9999
Subject: (usr-tc) MPIP in HyperArc
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-25 08:50:00
We currently have a TC hub with a Netserver and quads. We are looking er and quads. We are looking to expand and looking at the new Hyper bundles which have the HyperArcs and HDMs. The concern I have is the lack of MPIP support in the HyperArcs. Do we know when this will be available ? Jeff Binkley ASA Networ
Subject: Re: (usr-tc) Bug Reports and tech support
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-25 09:36:58
At 11:12 PM 2/24/98 -0500, you wrote: >Thus spake Jay Nitikman >>My only real complaint about USR tech support is that when I have a >>problem with my T/C they ask me to drop my firewall so that they can log >>into the hub and fix it. They have solved my problems this way but I >>have not learned how they fixed it, so the next time a similar problem >>occurs I have to call them backup. > >Amen, amen, and amen. I'm about to start refusing to give the techs any >access and just tell them that if they can't help me solve it then they >need to escalate the call to someone who can. I would have to differ in opinion here.. I do agree that the customer needs a complete detailed explanation of that procedures were used to remedy a situation. BUT, being the blind man on the phone and having someone read you a screen is very inefficient and can lead to mistakes.. How often do we say one thing and the other person understands it in a completely different way? My suggestion to you is to make sure you insist that the tech give you the gory details of what they are doing.. It is in our best interest to educate the customer to prevent repeat calls on the same issue. -m
Subject: Re: (usr-tc) Sample quad configuration..
From: Francis Fong <francisf@synergy-hk.com.hk>
Date: 1998-02-25 09:39:58
Something more! I do want sample configuration which can take 1200 & 2400 bps for Quad Modem, probably running the lastest firmware. Pls advise as we are facing problems of dial-in TCENH using 1200 or 2400 bps modem, at a 10-20% failure rate. Rgds, Francis On Tue, 24 Feb 1998, System Administrator wrote: > On Tue, 24 Feb 1998, Pete Ashdown wrote: > > > System Administrator said once upon a time: > > > > > >Can someone post a sample quad configuration that will accept ISDN calls? > > >I tried setting my ISDN call control options to the *exact* same thing as > > >my HDM ISDN call control, and now ISDN can't connect to the quads (yes, > > >I've got ISDN-GW set to 0). I'm trying to get better performance from the > > >quads than I do from the Munich, and not having much luck. > > > > What version are your Quads on, and what are those settings specifically? > > Currently, 5.6.7 (have the upgrade, haven't done 'em yet). > > I *did* get ISDN to connect, but I'm having a horrid problem (had this > problem before as well). For some reason, when I put ISDN calls on to the > quads, latency in general goes through the roof -- but ONLY for ISDN > connections. Not talking about UDP latency here, either. For example, my > average ping time across ISDN on quads is 100ms, instead of 50 or so when > I'm on the Munich. Telneting becomes painful and "spurty", whereas it's > perfectly smooth on the Munich. Exact same thing happens to ISDN calls on > HDMs. > > I keep hearing on this list how it's tres important to get your ISDN on to > quads, in terms of improving Quake/UDP latency issues, and I would *LOVE* > to do so if it weren't for this horrid overall latency problem. > > As a footnote... my solution to ISDN on the quads was to *ENABLE* > Async/Sync PPP conversion. However, enabling this on HDMs seems to have > the opposite effect (ISDN calls don't work). Very odd. > > For reference my Quad ISDN Call Control config is: > > V110 Rate Adaption : disable > Force Fixed Network Rate : notForced > Force Network Rate Speed : kbps56 > Enable 45-65 Second Link Delay : noDelay > Analog Calls Over Digital : enable > Async PPP/Sync PPP Conversion : enable > X.75 : disable > Set Date Mode of Modem : autodetect > Set Originate HDLC Protocol : none > Set Originate non-HDLC Protocol: none > Set Originate Analog Modem/Fax : analogModemFax > V120 : disable > X75 Frame Size : 2048 > X75 Window Size : 2 > > Regards, > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Bug Reports and tech support
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-25 11:11:38
Thus spake Mike >I would have to differ in opinion here.. I do agree that the customer needs >a complete detailed explanation of that procedures were used to remedy a >situation. BUT, being the blind man on the phone and having someone read you >a screen is very inefficient and can lead to mistakes.. How often do we >say one thing and the other person understands it in a completely different >way? I can understand that viewpoint, but lets consider this. I've never had *any* other tech support ask to get access to my boxes, and to be honest with you, up until at least very recently, I think I've known more about this equipment that the tech's that I'm talking to, so I really don't feel comfortable giving them access to my systems. Now, I could go for a compromise...I'll give the tech the public (read-only) community string, but the private stays with me...that will give them the ability to look in there and see what the current status is, but prevents them from being able to muck anything up. Yes, I'm paranoid about security, but is that really a "Bad Thing(tm)"? >My suggestion to you is to make sure you insist that the tech give you the >gory details of what they are doing.. It is in our best interest to >educate the customer to prevent repeat calls on the same issue. I would agree wholeheartedly with this, but to be honest, like I said, I just plain don't trust your people at this point. Last time I opened a ticket, the person asked me for the private community string and I told them that I didn't want them changing anything, and they said that they wouldn't, but then didn't have a good answer when I asked them why they needed the private community string if they weren't going to change anything. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) exitBeforeX2Connect(14)
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-25 11:16:54
Rick Payne said once upon a time: > John> I have a couple of hiper dsp cards in my chassis, and when i > John> do a performance report, under the x2status I get > John> exitBeforeX2Connect(14) does this mean that the modem > John> training time was too quick for this connection to establis > John> x2 status, and if so how di i increase the training time? > >I find modems will report "exitBeforeX2Connect" anytime a call started >out as X2 and then trained back to v34. I have this happen with callers who could connect with X2 on the Quads no problem. Now they can't establish X2 to the HDM, whereas others can. A bug?
Subject: Re: (usr-tc) archives
From: jpayne <scat@pcpartner.net>
Date: 1998-02-25 12:04:15
Thus spaketh (I've been waiting for this) Jeff Mcadams in part: >Someone was keeping an html'ized version somewhere I believe, but >don't know where off the top of my head. Just signed up to this list today - try this address in searchable form at: http://usr-tc.datasys.net Jeff Payne Siemens Telecom Networks jeff.payne@stn.siemens.com
Subject: Re: (usr-tc) archives
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-25 12:14:22
Thus spake jpayne >Thus spaketh (I've been waiting for this) Jeff Mcadams in part: Tee hee...I love it! :) >>Someone was keeping an html'ized version somewhere I believe, but >>don't know where off the top of my head. >Just signed up to this list today - try this address in searchable form at: > http://usr-tc.datasys.net Yup, that would be it (someone else mentioned this the other day as well). -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) MPIP help
From: Henry Moats <nc0419@corp.netcom.com>
Date: 1998-02-25 12:17:34
Folks, I am having trouble getting MPIP to work. I followed the instructions in the 3.4 release notes and i get a > I0 us,ppp,hmoats - Netwrk In CONNECTING on my second netserver. it will not establish a MP link. any ideas?? thanks ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________|
Subject: Re: (usr-tc) 1200 and 2400 baud connections
From: jpayne <scat@pcpartner.net>
Date: 1998-02-25 12:19:02
> From: John A Timon <timon@execulink.com> > > Hi all, I am having a problem with connecting. I have 3 chassis, 2 > with 12 quads, and 1 with 12 quads and 2 hiper dsp cards. all three > with i dentical configurations. The problem is on one of the chassis > I have a quad that is continually connecting at 1200 baud, I have > tried reseting the modem to default and writing to nvram and software > rebooting the card to no avail, am I missing something? Also i have > another chasis that every modem in it connects at 2400 baud > everytime. it has the exact configuration as the others, and i have > tried restoring the default there too, to no avail. Check the value you have set for "Link Rate Speed Select (&N)". If you have Total Control Manager, (highly recommended), Go to Configure, then "Signal Control Settings". Change the parameter of "Link Rate Speed Select (&N)" to variable from 1200 or 2400... Jeff Payne Siemens Telecom Networks jeff.payne@stn.siemens.com
Subject: Re: (usr-tc) 1200 and 2400 baud connections
From: jpayne <scat@pcpartner.net>
Date: 1998-02-25 12:19:02
> From: John A Timon <timon@execulink.com> > > Hi all, I am having a problem with connecting. I have 3 chassis, 2 > with 12 quads, and 1 with 12 quads and 2 hiper dsp cards. all three > with i dentical configurations. The problem is on one of the chassis > I have a quad that is continually connecting at 1200 baud, I have > tried reseting the modem to default and writing to nvram and software > rebooting the card to no avail, am I missing something? Also i have > another chasis that every modem in it connects at 2400 baud > everytime. it has the exact configuration as the others, and i have > tried restoring the default there too, to no avail. Check the value you have set for "Link Rate Speed Select (&N)". If you have Total Control Manager, (highly recommended), Go to Configure, then "Signal Control Settings". Change the parameter of "Link Rate Speed Select (&N)" to variable from 1200 or 2400... Jeff Payne Siemens Telecom Networks jeff.payne@stn.siemens.com
Subject: (usr-tc) exitBeforeX2Connect(14)
From: John A Timon <timon@execulink.com>
Date: 1998-02-25 12:37:24
First off, I have found this mailing list extremely helpful, and I appreciate all the responses I have recieved to my questions over the last couple of days. having said that, I have yet another question.... I have a couple of hiper dsp cards in my chassis, and when i do a performance report, under the x2status I get exitBeforeX2Connect(14) does this mean that the modem training time was too quick for this connection to establis x2 status, and if so how di i increase the training time? tia John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: Re: (usr-tc) Q: Performance on Chan T1
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-25 13:26:19
> While I'm on the subject: I've been told by a USR engineer that you > will seldom (if ever) see the upstream portion of an X2 connection go > over 28.8. Is this true? Anyone have any thoughts on why it might > be so? > I have a Courier 20 mhz, and I get 50666-52000/31200-33600 initial speeds. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) Sample quad configuration..
From: Greg Coffey <greg@coffey.com>
Date: 1998-02-25 13:36:27
Can someone post a sample quad configuration that will accept non-ISDN calls? We have a couple of channelized T1's and I'm looking for settings for error correction and compression in particular. Thanks, Greg Coffey, CoffeyNet ** Casper- Douglas USR x2 56k access ** 142 S. Center St. Wheatland, Pinedale, Lander, Lusk Casper, WY 82601 Douglas & Rawlins (307) 234-5443 http://www.coffey.com Open 8-6 M-F / 10-2 Saturday
Subject: Re: (usr-tc) new v.90 modem problem
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-25 13:40:02
On Wed, 25 Feb 1998, Richard Gamberg wrote: > There is at least *ONE* new USR v.90 modem in the field.... > And the owner can no longer get x2 speeds on local call. > > ISP is Prodigy Internet. Access is on Split Rock POP. > Who is Split Rock? > Anybody know if they aren't using the latest quad-modem firmware? Or is this > a show-stopping new bug? > > See > http://pages.prodigy.net/bbhi/x2-4.htm > Split Rock is a fairly large provider of pools for services like Prodigy. There firmware is problably just out of date. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) exitBeforeX2Connect(14)
From: John A Timon <timon@execulink.com>
Date: 1998-02-25 13:49:07
> >>>>> "John" == John A Timon <timon@execulink.com> writes: > > John> I have a couple of hiper dsp cards in my chassis, and when > i John> do a performance report, under the x2status I get John> > exitBeforeX2Connect(14) does this mean that the modem John> > training time was too quick for this connection to establis > John> x2 status, and if so how di i increase the training time? > > I find modems will report "exitBeforeX2Connect" anytime a call > started out as X2 and then trained back to v34. can i set it to prevent this retarining back to v34? John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: Re: (usr-tc) MPIP help
From: Henry Moats <nc0419@corp.netcom.com>
Date: 1998-02-25 14:34:55
On Wed, 25 Feb 1998, Brian wrote: > On Wed, 25 Feb 1998, Henry Moats wrote: > > > > > Folks, > > > > I am having trouble getting MPIP to work. I followed the instructions > > in the 3.4 release notes and i get a > > > > > I0 us,ppp,hmoats - Netwrk In CONNECTING > > > > on my second netserver. > > it will not establish a MP link. > > > > any ideas?? > > The link is established, more than likely. The MPIP in 3.4 shows up like > that for the second channel. Take a look at your ppp bundles or check the > client end to be sure. > > Brian > > do you see this same result in 3.6.28. reason: i upgraded from 3.5.34 to 3.6.28 and still see the same result. however, i see input and ouput on the client netserver. Hank > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > /-------------------------- signal@shreve.net -----------------------------\ > | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | > | Network Administrator | Perl, Linux | Web hosting, online stores, | > | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | > | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | > | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | > \-------------------------- 318-222-2638 x109 -----------------------------/ > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-25 14:52:08
Looking for expirences anyone else is having with fully loaded Hiper Boxes... We use Channelized T1 lines on ours, so we get 336 ports per box. We've been having some wierd results ranging from our own configuration problems to wierd telco stuff. Just curious if anyone out there has even tried yet?? | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: (usr-tc) EdgeServer Card
From: Timothy A. Deem <tdeem2@comsource.net>
Date: 1998-02-25 15:02:58
Anyone using the EdgeServer Pro? If you would, elaborate on what specific functions you're using it for and how generally it's performing for you.... I'm considering it for a project and wanted to hear from the "Real World" ;) Thanks, Timothy
Subject: (usr-tc) Edgeserver and Secure ID with Dialback
From: Ralph Helfenberger <rhelfenberger@comlight.ch>
Date: 1998-02-25 15:15:11
Hi Does anybody uses the Edgeserver together with SecureDynamics Secure ID in Dialback mode? We tried to establish this configuration but although the call is coming back from the Edgeserver there is never a connection on the PPP level. The same configuration works fine if only RAS dialback is used. Anybody any ideas? Thanks Ralph -- ========================================================================== R. Helfenberger Internet rHelfenberger@comlight.ch Comlight AG Tel +41 31 740 40 40 Tennisweg 21 Fax +41 31 740 40 90 3178 Boesingen Switzerland ==========================================================================
Subject: Re: (usr-tc) MPIP help
From: Brian <signal@shreve.net>
Date: 1998-02-25 15:18:04
On Wed, 25 Feb 1998, Henry Moats wrote: > > Folks, > > I am having trouble getting MPIP to work. I followed the instructions > in the 3.4 release notes and i get a > > > I0 us,ppp,hmoats - Netwrk In CONNECTING > > on my second netserver. > it will not establish a MP link. > > any ideas?? The link is established, more than likely. The MPIP in 3.4 shows up like that for the second channel. Take a look at your ppp bundles or check the client end to be sure. Brian > > thanks > > ______________________________________________________________________ > | > Henry Moats Network Services Support nc0419 ext 3671 | > ______________________________________________________________________| > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Brian <signal@shreve.net>
Date: 1998-02-25 15:19:26
On Wed, 25 Feb 1998, Curt Shambeau wrote: > Looking for expirences anyone else is having with fully loaded > Hiper Boxes... We use Channelized T1 lines on ours, so we get 336 ports > per box. We've been having some wierd results ranging from our own > configuration problems to wierd telco stuff. > > Just curious if anyone out there has even tried yet?? We only have a half full HiPer box, but I can tell you I have seen the ARC lose its interfaces before. Now I have all cards set up as static assignments and haven't had any problems since. And yes nmc discovery is on. Brian > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) MPIP help
From: Brent Jay <bjay@ionet.net>
Date: 1998-02-25 15:51:56
On Wed, 25 Feb 1998, Henry Moats wrote: > > Folks, > > I am having trouble getting MPIP to work. I followed the instructions > in the 3.4 release notes and i get a > > > I0 us,ppp,hmoats - Netwrk In CONNECTING > > on my second netserver. > it will not establish a MP link. > > any ideas?? > Make sure you have are using a good NTP server setup. Here are the steps you need to go through when initially setting up MPIP: nas1>set mpip on nas1>add mpipcl 199.199.199.1 secret nas1>add mpipcl 199.199.199.2 secret nas1>add mpipserv 199.199.199.1 secret nas1>save all nas2>set mpipserv 1 199.199.199.1 secret nas2>save all NAS1 is the server, and you would just add the clients you wanted and then set their mpipserver to the server address. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::
Subject: Re: (usr-tc) IP Pool BS: members=num_of_interfaces+1?
From: FBSD <fbsd@typhoon.co.jp>
Date: 1998-02-25 15:55:03
On Tue, 24 Feb 1998, Ricky Beam wrote: > Andrew Aken was heard to say: > >BTW, USR told me that you need to set the IP pools to 15% more IP > >addresses than the system needs to service the modems in your hub. > >Something to do with the time needed to release the IP addresses back > >into the pool upon disconnection. > > What the f*ck is this BS? The instant the PPP/SLIP/whatever session is > dropped, the address should be returned to the free address pool. What > kind of stupid programming BS have they done? What did they, write the > NAS firmware in JAVA? We use a NETServer/16, i.e. there are 16 interfaces. If we set the number of IP pool members to 16, we get an error of "no IP address available for dynamic address assignment," when the 16th user tries to log on. We have since set the number of pool memebers to 17 and things seem OK for now. Is it _official_ that there the number of IP pool members must be larger than the total number of interfaces? Is this documented anywhere? fbsd.
Subject: Re: (usr-tc) TCM Woes
From: Andrew Aken <ajaken@globaleyes.net>
Date: 1998-02-25 16:31:27
Steven, Look for an INI file in your boot directory (e.g. \WINNT). CyberGate Field Engineer wrote: > > To my dismay, Windows 95 was replaced with Windows NT workstation on my > PC. I inadvertently installed TCM ver. 5.0.2 instead of 5.1.1. > Apparently 5.0.3 does not use the "/usrsuite/tcm/devdb/" directory for > the database. I foudn this odd and just made 1 entry, saved it and was > able to retrieve the chassis. I was UNABLE to located where it put the > entry. > > I attempted to uninstall TCM 5.0.2 and it failed. I manually deleted the > /usrsuite/ directory and did a "regedit", deleting any reference to "tcm". > I then installed TCM 5.1.1, and copied the files in /usrsuite/tcm/devdb/ > from another machine, however the program did not read these files, but > instead read the 1 entry I made in the previous version instead. Where it > retrieved this information from I do not know. > > The issue is that we have almost 100 chassis, and I really do not wish to > enter that information all over again. In the past if we wanted to set up > another computer w/ TCM, we copied the "devdb" directory over and all was > fine. Any ideas? Thanks! > -- ======================================================= =========== Andrew Aken - President ========= ====== GlobalEyes Communications, Inc. ====== =Southern Illinois' Fastest Connection to the Internet= ========== http://www.GlobalEyes.net ======== =======================================================
Subject: (usr-tc) TCM Woes
From: CyberGate Field Engineer <steven@gate.net>
Date: 1998-02-25 16:38:49
To my dismay, Windows 95 was replaced with Windows NT workstation on my PC. I inadvertently installed TCM ver. 5.0.2 instead of 5.1.1. Apparently 5.0.3 does not use the "/usrsuite/tcm/devdb/" directory for the database. I foudn this odd and just made 1 entry, saved it and was able to retrieve the chassis. I was UNABLE to located where it put the entry. I attempted to uninstall TCM 5.0.2 and it failed. I manually deleted the /usrsuite/ directory and did a "regedit", deleting any reference to "tcm". I then installed TCM 5.1.1, and copied the files in /usrsuite/tcm/devdb/ from another machine, however the program did not read these files, but instead read the 1 entry I made in the previous version instead. Where it retrieved this information from I do not know. The issue is that we have almost 100 chassis, and I really do not wish to enter that information all over again. In the past if we wanted to set up another computer w/ TCM, we copied the "devdb" directory over and all was fine. Any ideas? Thanks! =========================================================================== S t e v e n R. S h e p h e r d =========================================================================== CyberGate Internet Technologies | ICQ: 1412432 An ACSI Company | NetDudeFL @ EFnet Field Engineer | E-Mail: steven@gate.net (800)NET-GATE/(954)429-8065 | 9542595004@alphapage.airtouch.com ============================================================================
Subject: (usr-tc) Accounting package buffer size?
From: Stefan Virsik <virsik@metronet.de>
Date: 1998-02-25 17:07:12
Hi all, one of our primary radius servers (authentication+accounting) recently died. Well, we smarties configured a secondary authentication server which answered the requests but doesn't accept the accouting packages. Now to the funny part: After restarting the accounting server, it received about 50 packets from each NETServer with huge delay times: Acct-Delay-Time = 134893 Does that really mean that the netserver card actually buffered the packets for e.g. 134839/3600 = 37.5 hours? Why did only these old packets arrive and not the most recent ones? Why did they arrive in the order of _as_cending delay times? Thanx for your eyes, Stefan
Subject: Re: (usr-tc) exitBeforeX2Connect(14)
From: David Bolen <db3l@ans.net>
Date: 1998-02-25 17:40:20
"John A Timon" <timon@execulink.com> writes: > I have a couple of hiper dsp cards in my chassis, and when i do a > performance report, under the x2status I get exitBeforeX2Connect(14) > does this mean that the modem training time was too quick for this > connection to establis x2 status, and if so how di i increase the > training time? There's nothing to increase. That value means that you had a call from an x2 capable modem (so the server detected that the x2 negotation process started) but during the training the client modem fell out of x2 (and likely back to V.34). This is a client modem decision, made once it determines that the line over which the call is being made will not be capable of supporting x2. Why it makes that decision could have any number of reasons based on its analysis of the line conditions during x2 training. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: (usr-tc) exitBeforeX2Connect(14)
From: Rick Payne <rickp@corp.netcom.net.uk>
Date: 1998-02-25 17:43:19
>>>>> "John" == John A Timon <timon@execulink.com> writes: John> I have a couple of hiper dsp cards in my chassis, and when i John> do a performance report, under the x2status I get John> exitBeforeX2Connect(14) does this mean that the modem John> training time was too quick for this connection to establis John> x2 status, and if so how di i increase the training time? I find modems will report "exitBeforeX2Connect" anytime a call started out as X2 and then trained back to v34. Rick -- Rick Payne, Senior Network Admin | Taxation is about extracting the NETCOM Internet Ltd. | maximum amount of milk with the rickp@corp.netcom.net.uk | minimum of moo.
Subject: (usr-tc) exitBeforeX2Connect(14)
From: Rick Payne <rickp@corp.netcom.net.uk>
Date: 1998-02-25 17:43:19
>>>>> "John" == John A Timon <timon@execulink.com> writes: John> I have a couple of hiper dsp cards in my chassis, and when i John> do a performance report, under the x2status I get John> exitBeforeX2Connect(14) does this mean that the modem John> training time was too quick for this connection to establis John> x2 status, and if so how di i increase the training time? I find modems will report "exitBeforeX2Connect" anytime a call started out as X2 and then trained back to v34. Rick -- Rick Payne, Senior Network Admin | Taxation is about extracting the NETCOM Internet Ltd. | maximum amount of milk with the rickp@corp.netcom.net.uk | minimum of moo.
Subject: Re: (usr-tc) new v.90 modem problem
From: David Bolen <db3l@ans.net>
Date: 1998-02-25 17:51:24
"Richard Gamberg" <bbhi@prodigy.net> writes: > There is at least *ONE* new USR v.90 modem in the field.... > And the owner can no longer get x2 speeds on local call. That's strange, since I definitely know that the V.90 modems will fall back to x2 when calling into our production hubs. Now we're still running the 5.{5/6}.6 code not the later 5.7 or 5.8 updates, so it doesn't seem to necessarily be an outdated quad code issue (unless the ISP is using some pre-release stuff, which would definitely be a bad thing). Of course, the x2 code in the V.90 modems isn't necessarily _exactly_ the same as the previous code the user may have been running (kind of like getting an x2 flash update at the same time as V.90) so there may be slight differences in line analysis for x2 calls. However, in general I've seen it make the same connections as older x2-only modems I had. Now I have seen different x2 releases do different things on the same line, so perhaps the user was previously using an old x2 release (say from early 97) but a later x2 update (which he/she didn't install) has problems on the line. So it's really the change in x2 more so than the V.90 stuff. It's not typical behavior though, and not something I expect to be a major "bug" in the V.90 code base. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Q: Performance on Chan T1
From: David Bolen <db3l@ans.net>
Date: 1998-02-25 18:46:16
"Phil Freed" <phil@freed.com> writes: > We have found that users with 33.6 modems cannot achieve 33.6 when > they connect through a channelized T1 using AMI (as opposed to B8ZS). > We've been told that 33.6 connections are impossible under these > conditions because of the bandwidth lost to Robbed Bit Signalling. I don't have hard data on a wide scale broken by T1 circuits at this point (and probably couldn't distribute it if I did), so I hesitate to talk in generalities, but I'm not sure I would say impossible - probably opt for "very unlikely" instead. To be honest, I don't recall seeing a significant difference between AMI and B8ZS encoding - it's the channelized T1 in the first place that affects you the most - and yes, because it's an RBS circuit. Not all channelized T1s are created the same, and there's more to it than just the "last mile" of circuitry to your equipment. Some carriers provide CT1 circuits as line side circuits, which can have additional gear in the path. Or if you're getting a service that provides you with a local presence that is remotely away from your gear, the forwarding of your calls may traverse paths that steal other bits, leading to multiple robbed bits along the callers path. Even though it is a local call to the dialup user, the act of forwarding it through the phone network is almost like a long-distance call. In an interesting example of the forwarding case, I can get the same quality connection to a local call to a CT1 circuit as I can get to a PRI circuit one exchange over. The act of just going through the telco network to get 10 miles down the road (as opposed to in my own building) actually robs a bit from the signaling path, even though I'm calling a PRI span (presumably, if I made a digital call I'd force the phone network to route me differently to avoid that RBS path, but...) Also, all of this assumes that you are talking about E&M circuits. If you're talking ground start or loop start that's a whole different ball park. In almost all cases, that is going to involve analog copper wires between the telco switch and a channel bank, which in turn creates the channelized T1 circuit to your equipment, and it's definitely going to be line side circuits, not trunk side. Such a connection is going to, in general, produce lower quality connections. (There are exceptions, where the circuit still goes digitally right into the switch - most often with CLECs or CAPs, since they don't have the massive investment in channel banks - or just a darn good channel bank, but it's not the rule) For example, I just took a peek at three racks of ours (8 spans, 192 modems each), one of which was AMI and the other two B8ZS, two of which were E&M and the other ground start. Here's the breakdown of rates for those callers who were using V.34 (as opposed to x2, V.32, etc..) right at the moment I checked: 12K 14.4K 16.8K 19.2K 21.6K 24K 26.4K 28.8K 31.2K E&M AMI 7 5 13 35 51 13 E&M B8ZS 1 3 4 3 13 46 68 8 Ground B8ZS 3 1 8 13 31 67 14 So as you see, in this relatively unscientific "poll", the AMI circuits actually seem to be performing better in this case, and the Ground start circuit has a definite shift to lower rates from both E&M circuits. And yet all 24 of these circuits at the three sites are "channelized T1" circuits, if you're talking at the 10,000 foot level. Of course, I didn't happen to catch any 33.6K connections on these racks at the moment. However, over the past 24 hour period, the following calls _ended_ at 33.6K: 33.6K Total V.34 %33.6 E&M AMI 18 5591 .32 E&M B8ZS 24 8520 .28 Ground B8ZS 0 6269 --- I checked ending stats to be realistic for users that either upshifted to 33.6K or held 33.6K for the duration of their session. Only a fraction of a percent did it, but it's definitely non-zero in the E&M cases. Note though that absolutely nobody got it in the ground start case. > But we've also been told by folks with PM3s (well, at least one major > ISP) that they regularly get 33.6 connections under these > circumstances. > > Of course, connection speed isn't the most important thing, since it > can always be renegotiated. Still, it _is_ what most of our users > see, and it's what we get lots of support calls on. It's really hard to judge stuff like this on word of mouth. As you say, what you really want is true line stats to compare. And you don't want info from the CONNECT string - since the initial connection is fraught with comparison difficulties. Sometimes it doesn't represent the link rate opposed to the DTE rate, and even when it does, it's really just an estimate with V.34 modems. Different modems estimate differently - USR tends to start lower and climb, others tend to overestimate and fall. And then there's the question of type of channelized T1 circuit in this other major ISPs cases versus those you are comparing to, as described above. Are they truly apples to apples? I guess I'd be leery of the word "regularly" too. If you were one of the .2-.3% of the users who got 33.6K into my racks, you'd probably get it all the time, so if I asked you your comment would be "regularly". So there's also a question of the breadth of their user base that gets it, as opposed to how often a specific set of users get it. > While I'm on the subject: I've been told by a USR engineer that you > will seldom (if ever) see the upstream portion of an X2 connection go > over 28.8. Is this true? Anyone have any thoughts on why it might > be so? That's also true - unlikely, but not impossible. Although I'd put it as more likely then the previous CT1 discussion. First, although the upstream for x2 (I'm not absolutely certain about V.90, but I think it's the same) uses V.34, it actually maxes out at 31.2K, not 33.6K. So asking for 31.2K on the back channel is akin to asking for 33.6K on a downstream V.34 session - it's the absolute best. One of the problems that the 56K protocols run into is that they stress the downstream channel much more, packing far more data into the signal. This intensifies the signal which exacerbates any local echo on the line and makes the echo canceler in the modem work much harder, among other things. This can lead to more difficulties in utilizing the upstream path, and thus you'll often find your upstream path suffering slightly on an x2/V.90 call than what you normally would get on a V.34 call. For example, on my home line (which isn't great) I can get 26.4K down, 24K-26.4K up with V.34. But when using x2 (or even better V.90), although my downstream can settle in at 45K, my upstream generally slips a tick and ends up at 21.6K-24K. But if you've got a good line, a 31.2K upstream link for an x2/V.90 call is certainly possible. It's much more likely to end up at 28.8K though, even on good lines. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Q: Performance on Chan T1
From: David Bolen <db3l@ans.net>
Date: 1998-02-25 19:06:23
"Phil Freed" <phil@freed.com> writes: > We have found that users with 33.6 modems cannot achieve 33.6 when > they connect through a channelized T1 using AMI (as opposed to B8ZS). > (...) You know when writing that last response it seemed somewhat familiar, so I peeked back at some old mail and found a thread that I had sent in the past. It's got a little more data from an analysis back then (although interestingly enough the percentage is about the same for 33.6K), so I thought I'd throw it out again. (I'm curious - for stuff like this would people just prefer that I just include a subject header reference for search purposes and point at the archive? I hate to clutter the list with repeat sendings. Opinions welcome (private mail, not list mail, please)) -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/ - - - - - - - - - - - - - - - - - - - - - - - - - In-Reply-To: Your message of Thu, 5 Jun 1997 19:56:25 -0400 (EDT) Message-ID: <CMM.0.90.2.865557950.db3l@foo.ans.net> "Adam Wills (Global 2000)" <sysadmin@global2000.net> writes: > Long story short, they claim its near impossible to > get 33.6 or 53k speeds out of a USR Tc on Chan t1. Well, I wouldn't quite say near impossible, but the stats are certainly in the very low percentages. As a very quick sample set, I ran stats on a few hubs at one of my channelized T1 nodes for a recent 24 hour period. Out of 17248 calls that used v.34, I saw the following for the v.34+ (31.2K and 33.6K) rates: Transmit (Hub->User) Receive (User->Hub) Initial Final Initial Final ---------------------------------------------------- 31.2K 1218 7.1% 1173 6.8% 2534 14.7% 3770 21.9% 33.6K 153 0.9% 56 0.3% 38 0.2% 172 1.0% So the numbers are small, but the one piece of information I can't account for is how many of those 17248 calls came from modems that themselves don't actually support the 31.2/33.6 rates. Also, these stats are spread over all callers, so obviously if a particular caller is on a line where the quality is the best, the odds of that particular person getting the highest rate is much higher than the overall system percentage. To be fair, here's a sample comparison from a set of racks at a site using PRI (different geographic region, but same timeframe) covering a total of 15432 v.34 calls: Transmit (Hub->User) Receive (User->Hub) Initial Final Initial Final ------------------------------------------------------ 31.2K 2368 15.3% 2270 14.7% 3298 21.3% 3459 22.4% 33.6K 593 3.8% 526 3.4% 1331 8.6% 1770 11.5% It is interesting to note that at least for this period on these nodes, the transit side to the user generally fell back a bit while the receive side from the user increased quite a bit. But even in a nice PRI case I'm only seeing about 3% of my callers get the highest rate, although it is about a 10x factor over a T1 configuration. It is only the initial transmit (according to the above numbers) that represents what a user will see in the CONNECT message. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications \ Phone: (914) 789-5327 | / 100 Clearbrook Road, Elmsford, NY 10523 \ Fax: (914) 789-5310 \ \-----------------------------------------------------------------------/ In-Reply-To: Your message of Thu, 5 Jun 1997 20:09:47 -0400 (EDT) Message-ID: <CMM.0.90.2.865558250.db3l@foo.ans.net> "Adam Wills (Global 2000)" <sysadmin@global2000.net> writes: > We are now going to try to arrange for PRI's for the TC rack, even due to > the increased price between PRI and Chan T1, its going to be necessary. > One thing I havent done, as stupid as it sounds, is plug a POTS line into > one of hte A/D quad cards and see if that gets me a 33.6 :) Now if that > does, I think I would just cry (and I think it will work too). Oh, one other thing you might like to try - check to see what the transmit signal strength is on your digital modem cards. I believe the default may still be -11db depending on how the modem cards were initialized, but I believe the recommendation is to run those cards connected digitally at -13db. Having a signal that is a bit "too hot" can affect the ability of the receiving modem to deal with it and might affect your connectivity. Of course conversely if your lines have more attentuation somewhere along the way, a hotter signal might be able to cut through it. I reset my digital modems to 13db as a standard part of our download process. (This won't affect your x2 stuff, for x2 the modem has a preset signal strength independent of the normal configured value). You can query change this using the mdmLiTransmitLevel MIB object (it's in the "line interface" group, so if you're using TCM check for that). The value is a number representing negative db. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications \ Phone: (914) 789-5327 | / 100 Clearbrook Road, Elmsford, NY 10523 \ Fax: (914) 789-5310 \ \-----------------------------------------------------------------------/ Cc: usr-tc@mail.xmission.com In-Reply-To: Your message of Thu, 5 Jun 1997 20:33:29 -0500 (CDT) Message-ID: <CMM.0.90.2.865561163.db3l@foo.ans.net> Brian <signal@shreve.net> writes: > Mine are set to -11db so I guess that still is the default. Everything is > brought in on PRI. So I gather setting it to a bit colder signal will > improve V.34 connect speeds? I'm not sure that I've got that much solid empirical evidence that the -13db does that much better than the -11db, but I do know that every 2db or can represent a "tick" in a protocol if you're heading the wrong way. There have also been some cases where we were given circuits from a 4ESS rather than 5ESS, and because of that the attenuation along the path was different (since we were entering the network at a "higher" point) so we had to attenuate the signal a bit further, and did see noticeble results. Probably a good thing to try for your own case would be to store some stats (if you aren't already) on connection rates, then change the value and see what happens over a comparable period of time with the change. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications \ Phone: (914) 789-5327 | / 100 Clearbrook Road, Elmsford, NY 10523 \ Fax: (914) 789-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) MPIP help
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-25 20:23:21
Thus spake Henry Moats >On Wed, 25 Feb 1998, Brian wrote: >> The link is established, more than likely. The MPIP in 3.4 shows up like >> that for the second channel. Take a look at your ppp bundles or check the >> client end to be sure. >do you see this same result in 3.6.28. reason: i upgraded from >3.5.34 to 3.6.28 and still see the same result. however, i see >input and ouput on the client netserver. Yes, and on 3.7.24 as well. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: (usr-tc) MPIP in HyperArc
From: Jeff Binkley <jeff.binkley@asacomp.com>
Date: 1998-02-25 21:48:00
We currently have a TC hub with a Netserver and quads. We are looking to expand and looking at the new Hyper bundles which have the HyperArcs and HDMs. The concern I have is the lack of MPIP support in the HyperArcs. Do we know when this will be available ? Jeff Binkley ASA Network Computing CMPQwk 1.42 9999
Subject: Re: (usr-tc) MPIP in HyperArc
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-25 21:49:34
At 09:48 PM 2/25/98 -0500, Jeff Binkley wrote: > > >We currently have a TC hub with a Netserver and quads. We are >looking to expand and looking at the new Hyper bundles which have >the HyperArcs and HDMs. The concern I have is the lack of MPIP >support in the HyperArcs. Do we know when this will be available ? > I too am anxiously awaiting mpip for the arc. However, life without it is not nearly as bad as you might think depending on your setup. I thought it was going to be a nightmare in our mixed setup but it really isn't a problem for us. (3 netserver chassis - 1 hiper chassis) If you have mpip functioning on a group of netserver equipped boxes with a loaded HiperArc box at the tail of the huntgroup, then there is only one instance of "broken" mpip. That is the last avail channel of the netserver spans bonded to a channel on the arc spans. Without doing some 'David Bolen style' calculations/observations, I can only estimate that the chances of hitting this particular instance with 2B+D is somewhere between 1:100 and 1:1000. It just doesn't happen very often. We have few isdn customers compared to pots and our first few boxes say full round the clock anyway.. So it was nice for us to go ahead with our plans for hiper purchases (to reduce quake lag) and not be in too big a rush for mpip.. But I would like to hear a rumor, date, etc... allen
Subject: Re: (usr-tc) Version 3.02 - slow speeds?
From: Brian <signal@shreve.net>
Date: 1998-02-25 22:40:37
On Wed, 25 Feb 1998, mt wrote: > Joe: > > We run TCS 3.01 and have similiar problem. > I always use Microcom 28.8 to connect to the TC (48ports) modem . > My WIN95 dialup network PPP status showed connecting at 28.8K. > But when I switched to MS-DOS prompt and use FTP command > to connect a unix server that is in the same subnet as the NMC card, > my download (GET) speed is 1.2K/sec. After the download > I issued upload (PUT) the same file and I got 5.0K. > I tried the "GET" twice and resulte same. > Not the best way to test. There are alot of things to account for here, even as much as the cabling to that machine your were downloading from, flaky nic driver, etc. Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Version 3.02 - slow speeds?
From: mt <tsaim@mft.com>
Date: 1998-02-25 22:43:30
Joe: We run TCS 3.01 and have similiar problem. I always use Microcom 28.8 to connect to the TC (48ports) modem . My WIN95 dialup network PPP status showed connecting at 28.8K. But when I switched to MS-DOS prompt and use FTP command to connect a unix server that is in the same subnet as the NMC card, my download (GET) speed is 1.2K/sec. After the download I issued upload (PUT) the same file and I got 5.0K. I tried the "GET" twice and resulte same. See cut note below ftp> ls -la -rw-rw-r-- 1 ftp ftp 368640 Jan 29 02:28 db1.mdb ... 226 Transfer complete. 745 bytes received in 0.55 seconds (1.35 Kbytes/sec) ftp> get db1.mdb 200 PORT command successful. 150 Opening BINARY mode data connection for db1.mdb (368640 bytes). 226 Transfer complete. 368640 bytes received in 307.20 seconds (1.20 Kbytes/sec) ftp> put db1.mdb db2.mdb 200 PORT command successful. 150 Opening BINARY mode data connection for db2.mdb. 226 Transfer complete. 368640 bytes sent in 72.83 seconds (5.06 Kbytes/sec) ftp> get db1.mft test.mdb 200 PORT command successful. 550 db1.mft: No such file OR directory. ftp> get db1.mdb test1 200 PORT command successful. 150 Opening BINARY mode data connection for db1.mdb (368640 bytes). 226 Transfer complete. 368640 bytes received in 307.36 seconds (1.20 Kbytes/sec) I called USR/3COM support 3 times, (case 44195) , some one sugget I should upgrade my NMC from 8MB to 16MB RAM. But my doubt was the entire hub has only 2 users on when this kind of problem happened. And the problem was always there regardless how man users were there ? One other things bugs me, this TC reboot by iteself sometimes. The interval is about 1 to 3 days. If you or someone here has the solution, can you please help ? Thanks Meng tsaim@mft.com Joe Zulanas wrote: > At 01:52 PM 2/21/98 -0600, you wrote: > > > >Ever since we upgraded to Version 3.02 for the TCH (we have the 2059 > >bundle, quad digital modem cards, etc.), we have had a number of customers > >complain about slow connection speeds through the TC. Has anyone else had > >this problem? > > > > > > We have had this problem when they connect it is @ 26.4 or slightly less. > We also get the problem they will connect @ 44 and only get ~ 2 to 3 K > throughput even on the same internal network. > Has anybody seen this problem? > Joseph Zulanas > CEO > __________________________________________________________ > The Metropolitan Network, Inc To The > Internet & Beyond! > We supply you with all the Internet you can eat, for just $19.95. > __________________________________________________________ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) new v.90 modem problem
From: Robert Sanders <rsanders@mindspring.net>
Date: 1998-02-25 23:24:35
David Bolen <db3l@ans.net> writes: > That's strange, since I definitely know that the V.90 modems will fall > back to x2 when calling into our production hubs. Now we're still > running the 5.{5/6}.6 code not the later 5.7 or 5.8 updates, so it > doesn't seem to necessarily be an outdated quad code issue (unless the > ISP is using some pre-release stuff, which would definitely be a bad > thing). I'd bet that the Splitrock modem refusing to negotiate X2 is a Bay 5399/8000 RAC. We have one running X2 code that will fallback to V.34 if I call with a V.90 Sportster. If I disable V.90 on the Sportster (S32=66), it connects at X2. Go figure. -- Robert
Subject: Re: (usr-tc) new v.90 modem problem
From: David Bolen <db3l@ans.net>
Date: 1998-02-26 01:39:55
Robert Sanders <rsanders@mindspring.net> writes: > I'd bet that the Splitrock modem refusing to negotiate X2 is a Bay > 5399/8000 RAC. We have one running X2 code that will fallback to V.34 > if I call with a V.90 Sportster. If I disable V.90 on the Sportster > (S32=66), it connects at X2. Go figure. Interesting - I wonder how much of the code on the Bay is directly licensed from 3Com versus local implementation - maybe it's getting confused during the V.8 stuff for V.90 and just falling all the way back to V.34 without checking for x2. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Bug Reports and tech support
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-26 01:55:10
Hey, >They seem to be in a hurry to move to the next call. First people complain that they have to stay on hold forever and then they complain that the techs are working faster to get to all the calls. :) Sheesh, you can't have it both ways! Steve
Subject: Re: (usr-tc) Netserver idle-timeout
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-26 02:11:51
Richard: -----Original Message----- >Is it possible to disable someone from making more that one simulatenous ppp >connection?. I would think the best way to do this would be with your Radius software. There is likely a concurrent logins settings somewhere, depending on your software. Set it to 1 and Radius with fail to authenticate the user on the second login attempt. This assumes, of course, that your Radius authenticator also does your accounting and keeps track of what usernames are currently logged in. Steve
Subject: Re: (usr-tc) Q: Performance on Chan T1
From: David Bolen <db3l@ans.net>
Date: 1998-02-26 02:32:42
"Stephen W. Buza" <steve@nemaine.com> writes: > That's a lot of bits to rob. > > Thoughts? My bet is it's a second A/D conversion and not robbed bits along the path, but I'd need to see the output from an ATY11 command for each call to be sure. This is actually not as uncommon as you might think (long distance working while local doesn't) - certainly more than is desirable. Generally it occurs because the local telco routes local calls differently than long distance calls and the local routing involves an extra analog hop in the middle somewhere. My best guess as to why this happens is that the telco routes long distance calls over newer facilities to major interexchange points which are likely to be updated more regularly as they are central locations. Whereas local calls may go through older infrastructure within the telco, some of which may still be using analog hops between different switching centers. The other possibility would be some strange padding within the local call path that is slightly out of spec or something that wasn't identified during x2 development. In such a case, V.90 should fix this since it will dynamically build an appropriate constellation for the particular padding involved. Audibly to the user, this case would probably sound like an x2 connection that tries to succeed but ends up at V.34, as opposed to call where you never hear even an attempt at x2 but just V.34 (the latter being too much attenuation due to the dual A/D). -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Q: Performance on Chan T1
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-26 02:33:26
David / Phil: -----Original Message----- >"Phil Freed" <phil@freed.com> writes: > >> We have found that users with 33.6 modems cannot achieve 33.6 when >> they connect through a channelized T1 using AMI (as opposed to B8ZS). >> We've been told that 33.6 connections are impossible under these >> conditions because of the bandwidth lost to Robbed Bit Signalling. <snip> >In an interesting example of the forwarding case, I can get the same >quality connection to a local call to a CT1 circuit as I can get to a >PRI circuit one exchange over. The act of just going through the >telco network to get 10 miles down the road (as opposed to in my own >building) actually robs a bit from the signaling path, even though I'm >calling a PRI span (presumably, if I made a digital call I'd force the >phone network to route me differently to avoid that RBS path, but...) Here's an interesting example of a call forwarding case: I have a customer who dials his local virtual circuit with his x2 modem and can achieve no better than 26.4Kbps. If he dials my T/C rack directly he gets x2. This is a 20-mile long distance call to his west. He can even call another town, 40-miles to his east, hit a virtual call forwarding circuit there that traverses all 60 miles to the west to my T/C rack and get x2. But he can't get it at his local number. I assume that all three calls get routed through his local CO. That's a lot of bits to rob. Thoughts? Steve Buza
Subject: Re: (usr-tc) Netserver idle-timeout
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-26 03:13:29
Stephen W. Buza was heard to say: >Richard: >-----Original Message----- >From: Richard Bosire <bosire@nairobi.africaonline.co.ke> >To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> >Date: Wednesday, February 25, 1998 12:49 AM >Subject: Re: (usr-tc) Netserver idle-timeout > ... >This assumes, of course, that your Radius authenticator also does your >accounting >and keeps track of what usernames are currently logged in. AND that the RADIUS system can keep such information CORRECTLY. I can go look through my local user database and see lots of users showing dozens of connections when they aren't even logged in. --Ricky
Subject: Re: (usr-tc) Q: Performance on Chan T1
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-26 03:24:10
David, Thanks for the quick reply. I suggested that an additional analog hop might be the problem, but my provider's network manager doesn't see it as a possibility. Ho hum.. Worker bees don't know anything about gathering pollen. Steve -----Original Message----- >"Stephen W. Buza" <steve@nemaine.com> writes: > >> That's a lot of bits to rob. >> >> Thoughts? > >My bet is it's a second A/D conversion and not robbed bits along the >path, but I'd need to see the output from an ATY11 command for each >call to be sure. > >This is actually not as uncommon as you might think (long distance >working while local doesn't) - certainly more than is desirable. >Generally it occurs because the local telco routes local calls >differently than long distance calls and the local routing involves an >extra analog hop in the middle somewhere. > >My best guess as to why this happens is that the telco routes long >distance calls over newer facilities to major interexchange points >which are likely to be updated more regularly as they are central >locations. Whereas local calls may go through older infrastructure >within the telco, some of which may still be using analog hops between >different switching centers. > >The other possibility would be some strange padding within the local >call path that is slightly out of spec or something that wasn't >identified during x2 development. In such a case, V.90 should fix >this since it will dynamically build an appropriate constellation for >the particular padding involved. Audibly to the user, this case would >probably sound like an x2 connection that tries to succeed but ends up >at V.34, as opposed to call where you never hear even an attempt at x2 >but just V.34 (the latter being too much attenuation due to the dual A/D). > >-- David > >/-----------------------------------------------------------------------\ > \ David Bolen \ Internet: db3l@ans.net / > | ANS Communications, Inc. \ Phone: (914) 701-5327 | > / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ >\-----------------------------------------------------------------------/ > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Netserver idle-timeout
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-26 03:31:39
Nice software :) Mine works. Steve -----Original Message----- >Stephen W. Buza was heard to say: >>Richard: >>-----Original Message----- >>From: Richard Bosire <bosire@nairobi.africaonline.co.ke> >>To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> >>Date: Wednesday, February 25, 1998 12:49 AM >>Subject: Re: (usr-tc) Netserver idle-timeout >> >... >>This assumes, of course, that your Radius authenticator also does your >>accounting >>and keeps track of what usernames are currently logged in. > >AND that the RADIUS system can keep such information CORRECTLY. I can >go look through my local user database and see lots of users showing >dozens of connections when they aren't even logged in. > >--Ricky > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: (usr-tc) opinion needed - mass software upgrade.
From: Henry Moats <nc0419@corp.netcom.com>
Date: 1998-02-26 07:22:18
I'm looking for opinions or tips on how one would upgrade hundreds of chassis in a small time frame. possible answers that i thought of. write a program using snmp. write a program using tcmunix. purchase 3com software ( i heard they have something to do this, but cost quite a bit ) hire temps to manually do the task. any ideas??? thanks ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________|
Subject: Re: (usr-tc) Bug Reports and tech support
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-26 08:09:16
Thus spake Stephen W. Buza >>They seem to be in a hurry to move to the next call. >First people complain that they have to stay on hold forever >and then they complain that the techs are working faster to >get to all the calls. :) Sheesh, you can't have it both ways! The way to get hold times down is not to rush through calls, but to hire more people (which they've done) or train people better so they can handle problems more efficiently. IMO though, that shouldn't require getting access to the systems that are being asked about. :) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) MPIP in HyperArc
From: buster_joseph/mw/us/3com@usr.com
Date: 1998-02-26 08:59:39
--IMA.Boundary.816505888 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part MPIP will be available in the 4.1.x release of HiPer ARC. 4.1 will be part of TCS 3.5. Beta for TCS 3.5 will follow after we finish up TCS 3.1. Buster jeff.binkley@asacomp.com (Jeff Binkley) AT Internet on 02/25/98 08:50:00 AM cc: --IMA.Boundary.816505888 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part We currently have a TC hub with a Netserver and quads. We are looking er and quads. We are looking to expand and looking at the new Hyper bundles which have the HyperArcs and HDMs. The concern I have is the lack of MPIP support in the HyperArcs. Do we know when this will be available ? Jeff Binkley ASA Networ - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message. --IMA.Boundary.816505888 Content-Type: text/plain; charset=US-ASCII; name="RFC822 message headers" Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Content-Disposition: attachment; filename="RFC822 message headers" Received: from usr.com (mailgate.usr.com) by robogate2.usr.com with SMTP (IMA Internet Exchange 2.02 Enterprise) id 4F4A8060; Wed, 25 Feb 98 17:23:50 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id RAA02544; Wed, 25 Feb 1998 17:04:25 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0y7q3V-0002Nd-00; Wed, 25 Feb 1998 16:16:01 -0700 Received: from linux.asacomp.com [199.178.136.5] (root) by lists.xmission.com with esmtp (Exim 1.73 #4) id 0y7q3Q-0002Kh-00; Wed, 25 Feb 1998 16:15:57 -0700 Received: from bbs2.asacomp.com (bbs2.asacomp.com [199. 178.136.12]) by linux.asacomp.com (8.8.5/8.8.4) with SMTP id SAA25329 for <USR-TC@LISTS.XMISSION.COM>; Wed, 25 Feb 1998 18:20:32 -0500 Received: by asacomp.com (NetXpress 2.54.beta3) with SMTP id 9C6EFA; Wed, 25 Feb 1998 18:15:13 -0500 X-Mailer: NetXpress 2.54.beta3 Organization: ASA Network Computing Message-ID: <35.327094.7@asacomp.com> Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.816505888--
Subject: (usr-tc) forcing high speed connection
From: John A Timon <timon@execulink.com>
Date: 1998-02-26 09:38:10
Is there a way to force an incoming x2 connection to send to the user at say 52k and recieve at say 28.8k ? John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: Re: (usr-tc) new v.90 modem problem
From: Richard Gamberg <bbhi@prodigy.net>
Date: 1998-02-26 09:50:37
I relayed Robert's info to Dean who tried the S32=66, and he now gets a 54k connect to the Prodigy/SplitRock POP. So, it's not a show-stopper - just a mosquito. Thanks, Richard -----Original Message----- >David Bolen <db3l@ans.net> writes: > >> That's strange, since I definitely know that the V.90 modems will fall >> back to x2 when calling into our production hubs. Now we're still >> running the 5.{5/6}.6 code not the later 5.7 or 5.8 updates, so it >> doesn't seem to necessarily be an outdated quad code issue (unless the >> ISP is using some pre-release stuff, which would definitely be a bad >> thing). > >I'd bet that the Splitrock modem refusing to negotiate X2 is a Bay >5399/8000 RAC. We have one running X2 code that will fallback to V.34 >if I call with a V.90 Sportster. If I disable V.90 on the Sportster >(S32=66), it connects at X2. Go figure. > > -- Robert >
Subject: (usr-tc) need help with TCM and NMC card that was repaired
From: System Administrator <root@wingnet.net>
Date: 1998-02-26 09:53:44
We had an NMC card go bad. I sent it off for repair and eventually got it back. In the meantime, I used a spare NMC card in the chassis. I also upgraded the chassis to the then-current code (3.1?). The problem is, now that I've got the fixed NMC card back, it has the older code on it. I'm wondering what I need to do to get my old card updated. Is the TCM software backwards compatible so that I can stick this NMC with older code and then flash it through TCM? Also, I got a memory upgrade for the spare card (from 4 -> 16 meg). It appears from reading the notes that came with the upgrade that if I put the memory upgrade in, the card will be devoid of operating software. Is this correct? If so, what next? Thanks, WingNET System Administrator 423-559-LINK (v) 423-559-5444 (f)
Subject: Re: (usr-tc) need help with TCM and NMC card that was repaired
From: Jeff Mcadams <jeffm@iglou.com>
Date: 1998-02-26 10:09:13
Thus spake System Administrator >We had an NMC card go bad. I sent it off for repair and eventually >got it back. >In the meantime, I used a spare NMC card in the chassis. I also >upgraded the chassis to the then-current code (3.1?). >The problem is, now that I've got the fixed NMC card back, it has the >older code on it. >I'm wondering what I need to do to get my old card updated. Is the >TCM software backwards compatible so that I can stick this NMC with >older code and then flash it through TCM? Yes, just throw the new (old, but repaired really) card back in the chassis and pull it up, and TCM will find it and you can do a code upgrade then. >Also, I got a memory upgrade for the spare card (from 4 -> 16 meg). >It appears from reading the notes that came with the upgrade that if >I put the memory upgrade in, the card will be devoid of operating >software. Is this correct? If so, what next? Yes, the memory upgrade will leave the card totally devoid of operating system so you can't even contact it with TCM, you'll need a PC to plug into the console port and use PCSDL to push code into it. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456
Subject: Re: (usr-tc) IP Pool BS: members=num_of_interfaces+1?
From: jason_kelton@3com.com
Date: 1998-02-26 10:26:05
I have heard something previously that you need 1 more than is actually required. It maybe documented somewhere, but nowhere I know of.... Regards, Jason fbsd@typhoon.co.jp on 25/02/98 16:55:03 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/AU/3Com) On Tue, 24 Feb 1998, Ricky Beam wrote: > Andrew Aken was heard to say: > >BTW, USR told me that you need to set the IP pools to 15% more IP > >addresses than the system needs to service the modems in your hub. > >Something to do with the time needed to release the IP addresses back > >into the pool upon disconnection. > > What the f*ck is this BS? The instant the PPP/SLIP/whatever session is > dropped, the address should be returned to the free address pool. What > kind of stupid programming BS have they done? What did they, write the > NAS firmware in JAVA? We use a NETServer/16, i.e. there are 16 interfaces. If we set the number of IP pool members to 16, we get an error of "no IP address available for dynamic address assignment," when the 16th user tries to log on. We have since set the number of pool memebers to 17 and things seem OK for now. Is it _official_ that there the number of IP pool members must be larger than the total number of interfaces? Is this documented anywhere? fbsd. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) opinion needed - mass software upgrade.
From: Mark R. Lindsey <mark@vielle.datasys.net>
Date: 1998-02-26 10:34:13
Henry, : write a program using snmp. I'll definitely vote for this one -- a tftp server and a few basic operations are all it would take, and you can easily run and monitor dozens of simultaneous uploads at a time just using any-ol' Unix. A walk could be done to check chassis integrity after the uploads are performed, and you could just as easily automate retries or report generation. TCM is nice, but I'd imagine that trying to do such mass work with it (whether alone or with temps) would be like trying to rebuild an engine with a Crescent wrench. (I'll admit that I'm not familiar with what can be done with tcmunix insofar as scripting is concerned; I've yet to find the Linux release. <ahem>) Mark
Subject: Re: (usr-tc) Version 3.02 - slow speeds?
From: Lee Kuo <lee@cosmo.mitec.net>
Date: 1998-02-26 10:35:43
On Wed, 25 Feb 1998, Brian wrote: > Not the best way to test. There are alot of things to account for here, > even as much as the cabling to that machine your were downloading from, > flaky nic driver, etc. Version 3.02 definately was the problem. We have downgraded to the more-or-less-stable-but-buggy 2.5.1, and no longer have slow throughput.
Subject: (usr-tc) TCC Quads
From: Jose de Leon <jadiel@thevision.net>
Date: 1998-02-26 11:58:55
I have a Total Control Chassis with PRI and Quads. I've never used Quads before, are they like modems, do the Quad modems need to be configured with Init strings like other modems? I've just left them at default settings and they seem to be working find, but I just want to know if I should do anything else. Thanks, Jose
Subject: Re: (usr-tc) USR TC Racks for sale
From: Terry Kennedy <terry@olypen.com>
Date: 1998-02-26 11:59:47
Why are you selling these? -----Original Message----- > >We have 2 USR TC racks as follows: > >1 chassis >1 PRI card >12 Quad Modem Cards (48 modems total) >1 netserver card >1 nmc card >dual power supplies >fan tray >nmc memory upgrade >all cables manuals etc > >3 months old > >$8500.00 each > >We also have 3 bundles as follows: > >1 high density chassis w/ integrated fan (capible of 336 modems total) >2 HiPer DSP cards (24 modems per card, 48 modems total) >1 HiPer ARC card >1 NMC card with memory upgrade >dual power supplies > >these units are *new* > >$11500 > > >/-------------------------- signal@shreve.net -----------------------------\ >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | >| Network Administrator | Perl, Linux | Web hosting, online stores, | >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | >\-------------------------- 318-222-2638 x109 -----------------------------/ > > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Netserver idle-timeout
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-26 12:37:09
On Thu, 26 Feb 1998, Ricky Beam wrote: > Stephen W. Buza was heard to say: > >Richard: > >-----Original Message----- > >From: Richard Bosire <bosire@nairobi.africaonline.co.ke> > >To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > >Date: Wednesday, February 25, 1998 12:49 AM > >Subject: Re: (usr-tc) Netserver idle-timeout > > > ... > >This assumes, of course, that your Radius authenticator also does your > >accounting > >and keeps track of what usernames are currently logged in. > > AND that the RADIUS system can keep such information CORRECTLY. I can > go look through my local user database and see lots of users showing > dozens of connections when they aren't even logged in. > Never had that problem here (except from older livingston PM2s, which seem to have some sort of odd bug when it comes to retransmitting Start/Stop packets). What radius server are you using? Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-26 12:43:30
Brian said once upon a time: > >On Wed, 25 Feb 1998, Curt Shambeau wrote: > >> Looking for expirences anyone else is having with fully loaded >> Hiper Boxes... We use Channelized T1 lines on ours, so we get 336 ports >> per box. We've been having some wierd results ranging from our own >> configuration problems to wierd telco stuff. >> >> Just curious if anyone out there has even tried yet?? > >We only have a half full HiPer box, but I can tell you I have seen the ARC >lose its interfaces before. Now I have all cards set up as static >assignments and haven't had any problems since. And yes nmc discovery is >on. How did you do the static assignments? I wonder if this has any bearings on my "no login menu" problems.
Subject: Re: (usr-tc) USR TC Racks for sale
From: Ian Roy <iroy@smokies.amnix.com>
Date: 1998-02-26 13:04:59
Brian, I just saw your posting below. > > We have 2 USR TC racks as follows: > > 1 chassis > 1 PRI card > 12 Quad Modem Cards (48 modems total) > 1 netserver card > 1 nmc card > dual power supplies > fan tray > nmc memory upgrade > all cables manuals etc > > 3 months old > > $8500.00 each > Are the modem cards Analog/Digital (ie. with the Analog NIC's ) ? Do you have the X2 keys enabled ? Are the units ISDN capable ? Also, do you have the software installed on them in electronic form ? Which versions of the software are they running ? We would be very interested in one of these units Thanks Ian Roy iroy@amnix.com
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-26 13:25:53
Curt Shambeau said once upon a time: >Looks like we got all the bugs worked out... I'm looking at TCM right now >with a fully loaded box with 336 ports full, and hopefully 2 of them will >be full shortly. > >We are running 2 ARC's in each chassis. The chassis are fully loaded with >14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, >except for a RIP v2 aggragate route bug, which we worked around. Curt, are you still running with 70 AMP power supplies?
Subject: (usr-tc) USR TC Racks for sale
From: Brian <signal@shreve.net>
Date: 1998-02-26 13:53:31
We have 2 USR TC racks as follows: 1 chassis 1 PRI card 12 Quad Modem Cards (48 modems total) 1 netserver card 1 nmc card dual power supplies fan tray nmc memory upgrade all cables manuals etc 3 months old $8500.00 each We also have 3 bundles as follows: 1 high density chassis w/ integrated fan (capible of 336 modems total) 2 HiPer DSP cards (24 modems per card, 48 modems total) 1 HiPer ARC card 1 NMC card with memory upgrade dual power supplies these units are *new* $11500 /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: RE: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-26 14:09:54
> You running with one or two HiperARC's ? > I'm planning on two chassis (sp?) with two ARC's each and about 8 HDM's > each, rising to full capacity, as demand picks up. USR tech told me > performance limit was at 210 ports for an ARC (7 HDM's for us E1 users). > > We saw some problems with the modems going oper status down, and admin > status up. Strange, as it was always the 1st modem that got affected. An > other problem was the ARC Reboots, this was fixed with 4.0.72 code. Looks like we got all the bugs worked out... I'm looking at TCM right now with a fully loaded box with 336 ports full, and hopefully 2 of them will be full shortly. We are running 2 ARC's in each chassis. The chassis are fully loaded with 14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, except for a RIP v2 aggragate route bug, which we worked around. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-26 14:46:40
> >Looks like we got all the bugs worked out... I'm looking at TCM right now > >with a fully loaded box with 336 ports full, and hopefully 2 of them will > >be full shortly. > > > >We are running 2 ARC's in each chassis. The chassis are fully loaded with > >14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, > >except for a RIP v2 aggragate route bug, which we worked around. > > Curt, are you still running with 70 AMP power supplies? One of the fully loaded boxes are running 70AMP power supplies (dual). The other two are running the 130's (now that they are off backorder). I plan on converting the 70 AMP chassis over to 130 at some point, but it is working fine as is. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: (usr-tc) Analog multilink setup
From: Chuck Simons <clsimons@simons.net>
Date: 1998-02-26 16:20:14
Can any give me some hints or tips on setting up a multilink session with analog modems. After spending 3 hours on the phone with 3com, (And still waiting for a callback) I get the impression they have no idea what I am talking about. I've seen posts in the totalservice newsgroup about this subject. And was told by USR 8 months ago it could be done. Still this tech support group can't seem to understand what I'm talking about. All I want to do is, take 2 modems on 1 quad card and enable them for multilink. Set up the NT server for multilink and get a little more bandwidth. Anyone have any ideas? Chuck Simons..
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-26 16:23:12
Curt Shambeau said once upon a time: > >> > We are running 2 ARC's in each chassis. The chassis are fully loaded with >> > 14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, >> > except for a RIP v2 aggragate route bug, which we worked around. >> >> can you expound a little on this bug? > >If you have multiple IP pools assigned to an ARC, and try to have the >pools set aggregate, the first pool works fine, but the 2nd route never >gets advertised. We temporarily worked around it by turning off aggregate >on the 2nd IP pool. One thing I do on my pools, for both the ARC and the Netserver is to setup static routes in my core router for them, then I actually block the advertisements for those routes on the Cisco interface, so the Cisco doesn't spend time on processing them every 30 seconds.
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Brian <signal@shreve.net>
Date: 1998-02-26 16:38:39
On Thu, 26 Feb 1998, Pete Ashdown wrote: > Brian said once upon a time: > > > >On Wed, 25 Feb 1998, Curt Shambeau wrote: > > > >> Looking for expirences anyone else is having with fully loaded > >> Hiper Boxes... We use Channelized T1 lines on ours, so we get 336 ports > >> per box. We've been having some wierd results ranging from our own > >> configuration problems to wierd telco stuff. > >> > >> Just curious if anyone out there has even tried yet?? > > > >We only have a half full HiPer box, but I can tell you I have seen the ARC > >lose its interfaces before. Now I have all cards set up as static > >assignments and haven't had any problems since. And yes nmc discovery is > >on. > > How did you do the static assignments? I wonder if this has any bearings > on my "no login menu" problems. > set chassis slot 14 carD_TYPE hdm_24 owNER yes typE static ports 23 like that > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) USR TC Racks for sale
From: Brian <signal@shreve.net>
Date: 1998-02-26 16:39:55
On Thu, 26 Feb 1998, Terry Kennedy wrote: > Why are you selling these? We upgraded the bundles to newer HiPer bundles, so we don't need them anymore. The last 3 bundles we are selling below because we don't need them but got a good promotion pricing on these. Brian > -----Original Message----- > From: Brian <signal@shreve.net> > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Date: Thursday, February 26, 1998 11:56 AM > Subject: (usr-tc) USR TC Racks for sale > > > > > >We have 2 USR TC racks as follows: > > > >1 chassis > >1 PRI card > >12 Quad Modem Cards (48 modems total) > >1 netserver card > >1 nmc card > >dual power supplies > >fan tray > >nmc memory upgrade > >all cables manuals etc > > > >3 months old > > > >$8500.00 each > > > >We also have 3 bundles as follows: > > > >1 high density chassis w/ integrated fan (capible of 336 modems total) > >2 HiPer DSP cards (24 modems per card, 48 modems total) > >1 HiPer ARC card > >1 NMC card with memory upgrade > >dual power supplies > > > >these units are *new* > > > >$11500 > > > > > >/-------------------------- > signal@shreve.net -----------------------------\ > >| Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 > | > >| Network Administrator | Perl, Linux | Web hosting, online stores, > | > >| ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs > | > >| 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ > | > >| mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 > | > >\-------------------------- 318-222-2638 > x109 -----------------------------/ > > > > > > > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: RE: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Brian <signal@shreve.net>
Date: 1998-02-26 16:40:40
> > We are running 2 ARC's in each chassis. The chassis are fully loaded with > 14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, > except for a RIP v2 aggragate route bug, which we worked around. can you expound a little on this bug? > > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: RE: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-26 17:10:52
> > We are running 2 ARC's in each chassis. The chassis are fully loaded with > > 14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, > > except for a RIP v2 aggragate route bug, which we worked around. > > can you expound a little on this bug? If you have multiple IP pools assigned to an ARC, and try to have the pools set aggregate, the first pool works fine, but the 2nd route never gets advertised. We temporarily worked around it by turning off aggregate on the 2nd IP pool. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) opinion needed - mass software upgrade.
From: David Bolen <db3l@ans.net>
Date: 1998-02-26 18:55:53
Henry Moats <nc0419@corp.netcom.com> writes: > write a program using snmp. For what it's worth, this is the approach we've taken. While our normal management tool can handle downloads to individual hubs much as TCM does, we have a dedicated tool who's purpose in life is large scale parallel downloading. The actual downloading itself is fairly straight forward in terms of SNMP operations and one or two TFTP transmissions depending on card type. We've found that over time there are various wrinkles that can occur, and have ended up encoding most of our known "bug" hueristics into the tool to keep them in one place. (Stuff like trying downloads three times with a card reset between the second and third try to work around various transient download failures) I think directly coding for this particular sort of application is the only way to go for large scale code upgrade handling. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-26 19:12:47
On Thu, 26 Feb 1998, Pete Ashdown wrote: > Curt Shambeau said once upon a time: > > > >> > We are running 2 ARC's in each chassis. The chassis are fully loaded with > >> > 14 HDM cards, 2 ARCs and an NMC. Things are running pretty flawlessly, > >> > except for a RIP v2 aggragate route bug, which we worked around. > >> > >> can you expound a little on this bug? > > > >If you have multiple IP pools assigned to an ARC, and try to have the > >pools set aggregate, the first pool works fine, but the 2nd route never > >gets advertised. We temporarily worked around it by turning off aggregate > >on the 2nd IP pool. > > One thing I do on my pools, for both the ARC and the Netserver is to setup > static routes in my core router for them, then I actually block the > advertisements for those routes on the Cisco interface, so the Cisco > doesn't spend time on processing them every 30 seconds. Same here. Then you redistribute these static routes into your OSPF domain, and your whole network knows where everyone is at any one time. :) Hell of a lot easier than screwing with USR's dippy RIPv2 aggregation. Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: David Bolen <db3l@ans.net>
Date: 1998-02-26 19:22:01
System Administrator <sysadmin@evcom.net> writes: > Same here. Then you redistribute these static routes into your OSPF > domain, and your whole network knows where everyone is at any one time. :) > Hell of a lot easier than screwing with USR's dippy RIPv2 aggregation. That's more or less how we've set things up all along, and using RIPv1 (host announcements) has been just fine with us so far. :-) The local router(s) - Ciscos in our case - at each site know the overall dialup prefix(es) they are responsible for and are configured during installation to statically annouce them to the core. Those site routers listen for RIPv1 announcements, but only to get the next hop right to the appropriate NETServer, they don't get redistributed at all. The static announcements in the core also prevent any flapping of the dialup space within the backbone. The core routers in turn aggregate the prefixes for all the dialup sites into that core before forwarding it on to the rest of the backbone and eventually aggregated into single blocks for external announcement. A single site router is generally responsible for no more than 16 or so NETServers, so it's local routing table and/or overhead for routing is reasonably sized, even with RIPv1 with host routes. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/
Subject: RE: (usr-tc) Full Loaded Hiper Boxen... Anybody else??
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-26 20:14:11
You running with one or two HiperARC's ? I'm planning on two chassis (sp?) with two ARC's each and about 8 HDM's each, rising to full capacity, as demand picks up. USR tech told me performance limit was at 210 ports for an ARC (7 HDM's for us E1 = users). We saw some problems with the modems going oper status down, and admin status up. Strange, as it was always the 1st modem that got affected. = An other problem was the ARC Reboots, this was fixed with 4.0.72 code. Robert von Bismarck Assistant Network Manager Petrel Communications S.A. -----Original Message----- From: Curt Shambeau [SMTP:curt@execpc.com] Sent: mercredi, 25. f=E9vrier 1998 21:52 To: USR Total Control Mailing list Subject: (usr-tc) Full Loaded Hiper Boxen... Anybody else?? Looking for expirences anyone else is having with fully loaded Hiper Boxes... We use Channelized T1 lines on ours, so we get 336 ports per box. We've been having some wierd results ranging from our own configuration problems to wierd telco stuff. Just curious if anyone out there has even tried yet?? -- | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. | -- - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: RE: (usr-tc) JAVA Based TC Manager
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-26 20:24:19
Have you ever tried using MRTG for this kind of stuff, I use this for about every NAS I have (Livingston PM-2 and USR TC) and it works beautifully. I do the quesries via snmpget and querying the interfaces for PPP and connect speed (blue line on the graph is #interfaces connected at > 33601bps) it's read from right to left. Here's an example of the output daily and weekly of a TC with HiperARC. -----Original Message----- From: System Administrator [SMTP:root@wingnet.net] Sent: samedi, 22. novembre 1997 00:06 To: Craig Nelson Cc: usr-tc@xmission.com Subject: Re: (usr-tc) JAVA Based TC Manager I'd certainly be interested in trying it out. We have two chassis units here. Having it be able to do a daily chart/graph of total line usage would also be nice. I was using pmwho and some unix scripts to count lines in use every five minutes until the new code broke pmwho. I'd then import that into Excel to make a daily chart. Good graphical view. A bit convoluted to get it done though. > I'm in the middle of creating a Chassis Manager in JAVA that should be able > to run on any machine with a JAVA VM. WingNET System Administrator 423-559-LINK (v) 423-559-5444 (f) - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) opinion needed - mass software upgrade.
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-26 21:40:47
At 10:34 AM 2/26/98 -0500, Mark R. Lindsey wrote: [snip] >(I'll admit that I'm not familiar with what can be done with tcmunix >insofar as scripting is concerned; I've yet to find the Linux release. ><ahem>) > I'd be happy with a release that works with Solaris 2.6.. allen
Subject: Re: (usr-tc) Netserver idle-timeout
From: Stephen W. Buza <steve@nemaine.com>
Date: 1998-02-26 21:50:53
Bosire: I'm probably not using the same RADIUS server that you are, so I don't know specific settings. My concurrent logins limitation is done via an option box in my new user template. Steve -----Original Message----- > > >Stephen W. Buza wrote: > >> Richard: >> >> -----Original Message----- >> From: Richard Bosire <bosire@nairobi.africaonline.co.ke> >> To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> >> Date: Wednesday, February 25, 1998 12:49 AM >> Subject: Re: (usr-tc) Netserver idle-timeout >> >> >Is it possible to disable someone from making more that one simulatenous >> ppp >> >connection?. >> >> I would think the best way to do this would be with your Radius software. >> There is likely a concurrent logins settings somewhere, depending on your >> software. Set it to 1 and Radius with fail to authenticate the user on the >> second >> login attempt. >> >> This assumes, of course, that your Radius authenticator also does your >> accounting >> and keeps track of what usernames are currently logged in. >> >> Steve > >Hi Steve .. > >Do you know what exactly ( files , dictionaries) that i have to change in >radius config .. > >cheers and thanx >bosire > >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the message. >> For information on digests or retrieving files and old messages send >> "help" to the same address. Do not use quotes in your message. > > > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Analog multilink setup
From: BECCOM <beccom@aol.com>
Date: 1998-02-26 22:31:41
The client that is dialing in must have the last Dial Up Networking from microsoft with ISDN support. Even though you don't use ISDN, you need it. After you install the latest DUN then you need to setup your second modem on the client. You will see that the new client of DUN has a device tab at the bottom. After you call the server and establish a connection, you will have the option of allowing the 2nd modem to dial and connect also. Make sure that the NT server has MultiLink PPP enabled. Hope this helps BECCOM@aol.com
Subject: Re: (usr-tc) Routing?
From: System Administrator <sysadmin@evcom.net>
Date: 1998-02-27 01:03:19
On Fri, 27 Feb 1998, Bob Purdon wrote: > > OK, I have a mildly interesting routing problem here that people here > might be able to help me with. > > Our network looks roughly like this: > > > rtr1 rtr2 border tc1 tc2 host1 host2 > | | | | | | | > +--------+----------+-----------+---------+-------+-------+--- > > rtr1 & rtr2 (Cisco 1003's) support only EIGRP and RIPv1. Router 'border' > is a 2501 and supports the usual compliment of routing protocols. tc1 & > tc2 are Total Control racks with NETserver cards, capable of RIPv1 & > RIPv2. host1..x are Unix boxes that can run anything gated will grok. > > I need to be able to have tc1 & tc2 broadcast subnet routes using RIPv2, > but the only classless protocol rtr1 & rtr2 support is EIGRP. I can > redistribute RIPv2 into EIGRP, but the routes are always propogated with > the border router as the next hop, which isn't what I want. > > Does anyone have any good ideas on how to achieve what I want, without > buying extra hardware, and without introducing sub-optimal routing? Yowzers. Doesn't look like there is any way to create "perfect" routing with your setup (at least not without money/upgrades). What you CAN do is keep things almost exactly as they are, and it's really not all that bad -- here's why (assuming that rtr1 and rtr2 are configured and listen to icmp redirects from your internal network). 1. On your border router run a decent interior routing protocol, such as OSPF or EIGRP. (I'd suggest OSPF myself, better compatiblity with future equipment) 2. Continue to have your tcs do ripv2, and configure gated to play nice on unix boxen. 3. Have your border router redistribute into the interior protocol that you choose. 4. Do NOT configure rtr1 or rtr2 to do any sort of interior routing. When they need to send a packet to something off of tc1 or tc2, they'll end up talking to your border router, but that's not really a terribly thing because the border router will go ahead and forward the packet through your network and bounce an ICMP redirect packet back to the offending source saying "hey, that's cool, I can handle this traffic, but in the future do yourself a favor and just send right to this gateway". 5. Now, depending on the intelligence of rtr1/rtr2, they should hang on to this "temporary" route for quite some time, generally until there has been no traffic through the temporary for x amount of time. Next time it happens, the whole process starts at #4 again. In summary, you end up routing a *very* few extra packets through your border router. It's not perfectly optimal, but the performance loss is so minimal, there really isn't a point to worrying about it. This is how we handle nearly ALL of the "stupid" workstations on our network (win95, etc), that can't learn any sort of sophisticated routing. They run up against a border router and get redirected. We also block ICMP redirects at "sensitive" points, so that someone doesn't get frisky and cause havoc (although, if a given tcp/ip stack isn't braindead enough to listen to obviously inappropriate icmp redirects AND your IPs can't be spoofed -- this really isn't a factor). Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *
Subject: Re: (usr-tc) opinion needed - mass software upgrade.
From: Pete Ashdown <pashdown@xmission.com>
Date: 1998-02-27 12:07:16
Allen Marsalis said once upon a time: > >At 10:34 AM 2/26/98 -0500, Mark R. Lindsey wrote: >[snip] >>(I'll admit that I'm not familiar with what can be done with tcmunix >>insofar as scripting is concerned; I've yet to find the Linux release. >><ahem>) >> > >I'd be happy with a release that works with Solaris 2.6.. I was going to ask this question earlier, but I presume the latest TCM does not work with 2.6?
Subject: (usr-tc) Routing?
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-27 12:56:41
OK, I have a mildly interesting routing problem here that people here might be able to help me with. Our network looks roughly like this: rtr1 rtr2 border tc1 tc2 host1 host2 | | | | | | | +--------+----------+-----------+---------+-------+-------+--- rtr1 & rtr2 (Cisco 1003's) support only EIGRP and RIPv1. Router 'border' is a 2501 and supports the usual compliment of routing protocols. tc1 & tc2 are Total Control racks with NETserver cards, capable of RIPv1 & RIPv2. host1..x are Unix boxes that can run anything gated will grok. I need to be able to have tc1 & tc2 broadcast subnet routes using RIPv2, but the only classless protocol rtr1 & rtr2 support is EIGRP. I can redistribute RIPv2 into EIGRP, but the routes are always propogated with the border router as the next hop, which isn't what I want. Does anyone have any good ideas on how to achieve what I want, without buying extra hardware, and without introducing sub-optimal routing? Cheers, Bob.
Subject: Re: (usr-tc) Reported slowness
From: Charles Hill <chill@ionet.net>
Date: 1998-02-27 13:32:39
On Fri, 27 Feb 1998, John A Timon wrote: > Hey all, We are having an interesting problem; in that > some of our users are reporting _very_ slow performance on their > connections, I was suspicious of their modems at first until i > experienced it myself. I am using an USR 56K sportster, and on this > last connection I noticed very slow transfers :( I am also getting > periodic busy signals (although the chassis is less than 1/3 > capacity) > > any ideas?? What version is your sportster code? Quad code? -CH
Subject: (usr-tc) Reported slowness
From: John A Timon <timon@execulink.com>
Date: 1998-02-27 14:16:50
Hey all, We are having an interesting problem; in that some of our users are reporting _very_ slow performance on their connections, I was suspicious of their modems at first until i experienced it myself. I am using an USR 56K sportster, and on this last connection I noticed very slow transfers :( I am also getting periodic busy signals (although the chassis is less than 1/3 capacity) any ideas?? tia John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: (usr-tc) real dumb question
From: John A Timon <timon@execulink.com>
Date: 1998-02-27 14:25:09
what does tcs stand for, and how can i check what release i am using? John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: Re: (usr-tc) Reported slowness
From: John A Timon <timon@execulink.com>
Date: 1998-02-27 14:41:35
> > > Hey all, We are having an interesting problem; in that > > some of our users are reporting _very_ slow performance on their > > connections, I was suspicious of their modems at first until i > > experienced it myself. I am using an USR 56K sportster, and on this > > last connection I noticed very slow transfers :( I am also getting > > periodic busy signals (although the chassis is less than 1/3 > > capacity) > > > > any ideas?? > > What version is your sportster code? Quad code? i have no idea how to tell the quads are supervisio 5.6.7 and datapump is 5.0.0 John Timon The early bird gets the worm but, the second mouse gets the cheese
Subject: (usr-tc) Radius Debug Improvements
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-27 14:56:02
Quick note: I made some enhancements on my radius debug tool. It now supports 3Com/USR VSA attributes and dumps more info about the packet. Get it from http://coredump.ae.usr.com/raddebug -m
Subject: (usr-tc) Dual HiPer ARC's
From: Brian <signal@shreve.net>
Date: 1998-02-27 15:07:37
Some questions regarding dual hiper arcs: I have a chassis with 6 HDM's a 1 ARC in it now. I am going to be adding 4 more HDM's and 1 more ARC to it. Do the two ARC's have to use seperate IP pools? Or can I delcare the same pool on both and have them share it? Does MPP work between ARC's within the same hub? Is there any concerns or caveats I need to know about using 2 ARC's in one hub? (that's not clearly expalined in the docs) Brian /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Dual HiPer ARC's
From: Mike <mwronski@coredump.ae.usr.com>
Date: 1998-02-27 16:08:48
At 03:07 PM 2/27/98 -0600, you wrote: >Do the two ARC's have to use seperate IP pools? Or can I delcare the same >pool on both and have them share it? 2 Pools Will be required. >Does MPP work between ARC's within the same hub? No. MPIP is not supported in the current ARC code. >Is there any concerns or caveats I need to know about using 2 ARC's in one >hub? (that's not clearly expalined in the docs) You must turn off Chassis awareness. This is done by typing 'dissable nmc chassis_awareness' at the ARC CLI. -m
Subject: (usr-tc) Vendor-Specific RADIUS attributes
From: I. Dwayne Koonce <dwayne@txcyber.com>
Date: 1998-02-27 16:42:46
Hello...I'm new to the list, so bear with me if this has been asked since the last message in the archive... We're currently running Ascend RADIUS on a Linux box for authentication & accounting for two TC's, two Ascend's, and a Portmaster. We've just decided to expand the information that we keep in our logs (in a MySQL database, courtesy of homegrown patch), but some of the information we want is sent by the TC's in the Vendor-Specific attribute, which Ascend radiusd doesn't support. I've tried Merit AAA 3.5.6, which appears to go out of its way to support vendor-specific attributes, but it logs them as an empty string and outputs errors like the following: Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific attribute Vendor-Specific=vUSR-000000c700000001 Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific attribute Vendor-Specific=vUSR-0000901900000000 ...which I see from the list archives that some others have as well. In fact, it looks like there's been quite a lot of discussion along these lines, but the threads appeared to mostly die out with no solution posted. So, my question is: Is anyone currently using these attributes successfully (with source-available software)? If so, then what? ____________________________________________________________________________ I. Dwayne Koonce E-mail: dwayne@txcyber.com ____________________________________________________________________________ "It's dangerous to be right when the government is wrong." --Voltaire
Subject: Re: (usr-tc) IP Pool BS: members=num_of_interfaces+1?
From: System Administrator <root@light.lightlink.com>
Date: 1998-02-27 16:52:13
> We use a NETServer/16, i.e. there are 16 interfaces. If we set the number > of IP pool members to 16, we get an error of "no IP address available for > dynamic address assignment," when the 16th user tries to log on. We have > since set the number of pool memebers to 17 and things seem OK for now. > Is it _official_ that there the number of IP pool members must be larger > than the total number of interfaces? Is this documented anywhere? I can't believe anyone follows the volume on this list. Anyhow on our 16-I if we assign 205.232.34.1 as the starting IP, when it is full it will have assigned up to 34.17 leaving 34.15 out. Homer
Subject: Re: (usr-tc) IP Pool BS: members=num_of_interfaces+1?
From: Ken Hunter, Aspiring Technologies <ken@aspire.net>
Date: 1998-02-27 17:25:27
If you are running the comos'ish code on your netserver, it will count the s0 as well as the other 16 ports when determining how many addresses to chew up. that note is from a member of tech support - so in your case, 17 sounds goodly - 9 for the netserver 8/i Ken :) At 04:52 PM 2/27/98 -0500, you wrote: >> We use a NETServer/16, i.e. there are 16 interfaces. If we set the number >> of IP pool members to 16, we get an error of "no IP address available for >> dynamic address assignment," when the 16th user tries to log on. We have >> since set the number of pool memebers to 17 and things seem OK for now. >> Is it _official_ that there the number of IP pool members must be larger >> than the total number of interfaces? Is this documented anywhere? > > I can't believe anyone follows the volume on this list. > > Anyhow on our 16-I if we assign 205.232.34.1 as the starting >IP, when it is full it will have assigned up to 34.17 leaving >34.15 out. > > Homer > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > - ************************************************************************ Web Hosting, E-mail addresses, DNS services, Dedicated connectivity. 33.6, 56k, ISDN Dialup coming soon to Prince William County VA. Ken Hunter Aspiring Technologies mailto:ken@aspire.net 9304 Troy Drive http://www.aspire.net Nokesville, Va 20181 ************************************************************************
Subject: RE: (usr-tc) JAVA Based TC Manager
From: Robert von Bismarck <rvb@petrel.ch>
Date: 1998-02-27 17:39:18
This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------ =_NextPart_000_01BD43A6.A33F7900 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Argh... I love Bill Gates, "drag and drop" should be called "drag and lost" in win NT... Here's the two files... =20 =20 Robert -----Original Message----- From: Randy Cosby [SMTP:dcosby@infowest.com] Sent: jeudi, 26. f=E9vrier 1998 20:48 To: 'rvb@petrel.ch' Subject: RE: (usr-tc) JAVA Based TC Manager the output didn't come through, but I'd be interested in seeing it. On Thursday, February 26, 1998 12:24 PM, Robert von Bismarck [SMTP:rvb@petrel.ch] wrote: > Have you ever tried using MRTG for this kind of stuff, I use this for > about every NAS I have (Livingston PM-2 and USR TC) and it works > beautifully. I do the quesries via snmpget and querying the interfaces > for PPP and connect speed (blue line on the graph is #interfaces > connected at > 33601bps) it's read from right to left. >=20 > Here's an example of the output daily and weekly of a TC with HiperARC. >=20 >=20 >=20 >=20 >=20 > -----Original Message----- > From: System Administrator [SMTP:root@wingnet.net] > Sent: samedi, 22. novembre 1997 00:06 > To: Craig Nelson > Cc: usr-tc@xmission.com > Subject: Re: (usr-tc) JAVA Based TC Manager >=20 > I'd certainly be interested in trying it out. We have two > chassis=20 > units here. Having it be able to do a daily chart/graph of > total=20 > line usage would also be nice. I was using pmwho and some unix=20 > scripts to count lines in use every five minutes until the new > code=20 > broke pmwho. I'd then import that into Excel to make a daily > chart. =20 > Good graphical view. A bit convoluted to get it done though. >=20 > > I'm in the middle of creating a Chassis Manager in JAVA that > should be able > > to run on any machine with a JAVA VM. > WingNET System Administrator > 423-559-LINK (v) > 423-559-5444 (f) >=20 > - > To unsubscribe to usr-tc, send an email to > "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages > send > "help" to the same address. Do not use quotes in your message. >=20 > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. ------ =_NextPart_000_01BD43A6.A33F7900 Content-Type: image/gif; name="lsnpm4_span_ch-day.gif" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="lsnpm4_span_ch-day.gif" R0lGODdh9AGHAOMAAPX19cLCwmRkZP8AAADMAAAAAABmAP8A/wAA/wAGBAQBEACoEEAHBAQAFAAL BQQBCCwAAAAA9AGHAEAE/jDISau9OOvNu/9gKI5kaZ5oqq5s676TIAF0bd94ru987//AoHBILBqP yKRyyWw6n9CoQBaI1grYa5a25XaHBa71OhaXw+cyAD1mGwc9N5AglEPtT7zzK5zO1IBkS3pNhEyG g2qISotJcHQ3BHSTkpU0lZiZZm2KnXFGflWBgY1rnpxpqairOo+WmbCwBbGWjKdWpTy5N6GfpkZo bMHCgmCbuLdRu0XLRM3GRwQDtJWz1Ne1usl3is8Avdq/o8CHOZOJqsjpyk+w09jw2N7mANnzQPc/ i97gOmH/447M0qcpWz1IsHzkU7gtT8M9l2i9i0fLGkVINc7tGHjtoEF9/g/rgKLiK8m/YDm2LMPC clYYAgODSbJGc6ZNmDezoPyFEqaZnUAB8vxJB+jPn0KNDl2qVOdRpk+bJo1KFeqaqUNxaqU5reZW lznDbhXrFctWomPTlq3n1GpTqjCxvnWLlCWRfv4CmoznkZpFbBlrSYo4uNiQwQlvsMEIcR23Ihcn XpysdXI9wtf+dszIozBbz4E/Qgt3l2TKlgtJUdRMOR5rxIxtaOpLqbLovrgvgzbMzrEQ19gkt7Y9 HN7r4rN/wd6cmPAli5ecjfzjryU5HHgAPvOK3Hh3TLSrffc+GzyS1KSZyPoufP34zO/hx6/4Xhc/ 03pxxT8+n//++QUR/mfebqPpB6A7B7qXoHgLMtiggOD5VARe+eWxGgEIYIhAhhpKsiFy/r0X4ngj thbYQEYAN96GLLJ4EzUTfRhLhhx6mCF/LspIzYc1vtjgXy7Co6N83d1HHSDDpORFEQM0aU2PNkJ5 jZQQCjlklC0WQCOVO14pI4ta2jgjl5W02OJNZnIYZJlptgmmm2uCyJ48D/o4JZmZSbklnioOp+eZ CPARBIUVDtKFXUG4gSg+iu1ih6M2LAqSFs1I6kw+81hah6AKHcppdcJ8mhelIkUKqWKleoGeDYQW 6uqrsMYq66y01morL/jdquuuvPbq66/AjtFqOZEuiZ2xwSar7LLM/jZL67DNlrLqqKz0Vu1j194A xw8G3RRabAKFROx504niLKriqJNtY+sWIu4Pj0TioImTtjsuLkaaq2S6YOy0EW+MomOvwOoWbG0U 0tQp4cC2+IZuaUceK1Sv0kJ2GUMOu5vxvQfLJkl7lIXYsG7gkrtxwBDre9q0qDSi0cU6IMYvtgbT 3LHNPiBIZHecKVayORHGBi7Lx56R778meUqIpvgQCPQrhUEXy8g1O3Qy1dHAqDBrHjvtUXjV0GY1 w0cjHexfuV0RXyTFeH3ey2NXzS4Uduos54GylTi12wCzHXO94U6Y68rAcvdjcTkfc85yt0Uk2zE3 x81OdyBjovfd/g8eR0/Q5Yknb3R+J1r2RkzbejmddS4GNuJcMPdR433PHTlmlG9tO3LfKsgXZTMX nTKo1vmq9+nDEd+a8SG/XDJCn5st99PJX1P5RchPVj31twOGWbGCR3zuGnUfeP2FqWdf8+vF4Xgn AtP0GGSYtKjJ5pV+ZX946qN/zwWUXJIZqIeY2FKHsMGjHs2CfjPSUI0KOCT44YlPatLR+OABvh2I KIBuqt1F4mQc+WUQg19aoAQxhCUODjAT7uPQARUYQh21qYQ1GlH+kOQpU3mjSQNAhA5LQq0e7st5 P/Sh74T4MCJyD4hDDKISk8jEIi7RiU08ohFjR0XIRbGKvZPi/hCgtaxQWdFsxEBZGOuFmkRBsSQq SdV2sAi8TAnEX2bUIsau8kVqjXGOd0Rj8ACXxSRwUX+ADKQgB0lIV/2xkIhMpCIXyUgdHLKRkIyk JCfpq0eeJ1RLExUlN8nJTgrSkswgQxqxA0d9oOaUqEylKlfJyla68pWwjKUsZ0nLWtrylgXIIS53 ycte+vKXZdzi4L6HHqKdUXY4k1wynbCtnC1PaBgxphyVSc1j+gCUvyrlMjXGMJO8C2tEiJffXmIe sW2Pj9vk2B1mSEyUPU+d1eTmOwmWA3F2jRYe4Q88kRnPOP5OkMX8pjevNtBuIsEVuftOH/cpz9lN 8wfYFF4d/huaTnpWFJz9hFeA5jWcOc4To3tgZ7EmFsdHKYmk/uShQxnKUotmFKSXmB75qAHTltbU nX0YpjVvFVAc/CyUBDWZQQNnhYSlTqBC/ahIbVi4iSLhHNJkY0E/OtWVWrASMq2SZdLTA0qkiqqX LBdpokrRvMCtqzDj6kWrutakdsxu6eML6EL3w4L4VDpBVSsQ/rhHXuUCNgfhTXOIClbCWvWmQdBa +cr5OAsmdCYeeylbr9iDR5K1EJDtKuciG4vLehaphl2r9FA3n0ig6GmrY1xsPms0se4rNUnJDrLq wLfWca42mzHbT93a1tBKllyjLRK9zik0jtB0sB49bHIH/qXTh+JDaVpgKjNuK6+d4XMOQ1PPFG06 WX7OIbVYtR9F7lm/xNQWnT3rrnMr21xVaXJWhisMQr5jjdAExnEFaqw5v6pc3saTIlmdIHkYe0G6 3tW+9DitMpYqBpTqanhro094UkQy8DG2efnt7XWUMF/LBYe+CeqLgGUGGtV5JkIJpetCGCyOyw7i PwuKHbhIfN1iCUYwc+WvhoEaM9y2JqskEu9FfOpj4tAOmvhl3nqdygMuungP4QPQBHscZe3dEzfG xbBvvbuEKgOYtPfLXG7BS6eu0Q6/z3Xtdk2JxTw2rU9hvm6RZ7pR3p0ItBsmAp1jAWQwL2jENAV0 ea28/uYdRBSvLUaWmwlywcWSGXOWSc6+duvfuS2oz8JdrJjLl94lG7q9zIpvggRNH01T8GtnfpmC d0xk6/7Yz6MWsvgeNNuceu9cEDb1n2Xd0UKr99FC0tILP9Y/PmklSMbGnq5jTV81A/J0Lowhr/sz 7akh2pnJcTVFEKgz+aGQfmeChQo3mOxMkJqEHl62LJytP1GLO348GrSVjhcPE8J5153LdmLPLYsh cZvPpGXgtvcEpQNGsDXhTne9Bfjt+f1PgaNm8Toz6Q1hw0mFyTYTuiPYJmGjkIUAnBK6NfSaBYJ8 TATfkMGjbW8W/kmr+7l4mor34Uw36H/xLqHIAWU5/gQmvHjgdjmVAsXu1vZulEDAIQ6ByfSmO/3p UI96LHUp9apb/eqo/Kcnt871rntdCYemGE9yMUayS3WIDtar2ZGiRh6Ha8VAfS8QVzJ2fy7ajm5P OxIX2j2VRSt4pSOcqkSHScLvtFHDCDziSdWvTMldpYR/vOAjfxLFU0olkr9K4Z+7eVPKgeyYt7zW v0760pu+9FNIvepXz/rWu/71sI+97GdP+9rb/va4z73ud8/73vv+98APvvCHT/zitx4GyE++8pfP /OY7//nQj770p0+BsJ/++tjPfrOsr/3ue//7s+I++MdP/vILC9TmT7/61+8E8Xt00Xdnv/znj0j3 /tsnjWnXDv33z/9C2h+MiUZHm7BGLsVl3FVpZWWAS9BMhZVn/bVlnuZI6KcsFZNXDshqGfZbEEgD DOh2OYZYv1aAOrZXE5gs2qSBF4iCHoiBIygEHVg03VJOF7Z3CgiCYdV3g9RTQ5WCNRiCNngE9tRp 2SaDtcBaO7iClNVkJXg2OMWCKfWAPJiAUsiBaYVm5qUVZyaCB7iBTUiCt+YsJ9iDCLiFUUiGSOiC VRhY5DQghKZXYsiFjCBxFGiG1waFZziFWriAaQgYcmWEDXiHZ4cD/wdfXfiGZZiHP3iIN4VQRwZr MwiIiOiDy+WFfodrTJaIkIiJdeiEhRgErqBa/uODXoaYiUlTdJYYiWOIinAoiamoBJ84Z7jjh3a4 iUn4aV/YGMFEShXXiXSYgaNIi7/oizUgHMCmcDW2iq3Iim54TaCmf9KReHqQJHgli5woinioicLI LQGmUJCngsBYCCzmjJtiKdkRhqTTFnORjlWhjnTRjuz4jusYj+4oj/A4j/ZYj29hVGVBFvyoFvT4 j/cIkPgokAIpeoK4hEK1KF9gkE/UkGjFi6qoiMookWNoVJBmbcmYkci4jOx1i6agd7Vijk8ViBS5 kSV5kqTYY9vIM934kBCJjZtiiqLEd7Kig49DaS85kSlpkjt5GB+zNU3gVSTZk5kohyaYk3eF/pM0 qZPf2IstWFaYgGnydow8JpRO+YREWQNONpQVIpKcoZTRURR4lpXZGJFECVedVR8fWEHcgmIGtpQa CQzhyJX5sQvKAzAyA5NY6Y1lqZdUFl6ctpZp+JeU4AZg+ZQ2ZZRMyGgXwzeA1ZLXyJR9KZlfBXA2 l1vLQxoB8paTGIyHZ4uVOJNPNlCLAFhus24WSJaIaZZNSV5R6YgsGSm7hRE0dV8RGJfMMJcgSSur dlVQAwnkdGR+aY1XiZQ86ZtoyWyvM5j35Rf7VZzEeYkS6JFddF6agWq3BTt0uZrD2ZmRCVyKdZFy dU4+ZVz4JGnR2Z1LBXcN1jfxpwuZ1RnU/mUciaWae+mZ9xmZwTWVavkc8lVd4DVh25mfh8BgpySX 7Slbu7kR2fAzHTEiiZVdXeZruNma3VlzmzaE/+manSWgh7mMH0qDt3mQ1JmQAXgodIk2qKY23Khb ITopL2qcKGmhe7GfcWUitZkRyINmIpo43kmgSuiRo3kINpE2jRgyEUqjzPmjrDmZCAiLr6ltUjZY uHMQUSWWRjd6DslTtdlh/FljDTpftfg3WTiiM+qkx/lYlpmhQ6am5jaegimEvRmWMcpGilk4fAFo 39I65PlURYGenMmk3QmZScBRyRk9ANKYetppcnBiogGoZkqhJBqaCZpNXtZoAfqcLala/rO5pU3K nVXlpRhaamzKqZlmmz22p/e0anVaHTJZqcIDYxTUEc1hlxWGXMrRqhz5qUDaM6+mbnyobLMaWJtD m/VFm+j5ofvwqqG2Z8oJqZy5OH/6iDwKltTIl62WljZKc8A6Hm2DG9aGXPLVqEo6qWNlqTDXaDjq oGEjNmO2p+CRZdUoqE81QStZqs86g6aabzhhZuYUopmHK0JKgbL6I9DDOmJAXRUWWYTKq50YH1J5 o/i6VUearqBxYzfmaXeKrgMWZxjZmNbzsdpzYfH5h+WarVIKmOI5a7S2hrEpGNBhpA0LmrtKRn0l Xc5wqUEWmyRjjBQLGCqKYGkKqili/rGjqq0TO6U5epnFqJ3RhYM1q0efIo1gULB/1rMRZrQEVlvX ymUQW23q2lEdu2szO52UCk/lKJ02tI9jYRZq0Y9sWxNuG7fTSrdta7d4mxOfMUpGISHwyLcECRWA yxRwGxZzGxY5VLhve7hv27h0y7h567g2AbmKm7eUexYVhBI3C1HN9Xm52XkLGbCyqbUrq7RJa1eC qWRqk5ryqTvDca8em6+y22tpI7pa2bmfGSuAxm9v2q2RxqNooLo8erLTVScRy62na7WzO67MWp33 trzUpmnOkWowwzdd66bCKhEg5rthi2+9tjDCVKImqLPbm7zdO2SnNqffiZykuiPz/hOlCXQ55da7 5ruzDSKAWkpMysuy9Vu+p0a8K1qx15Mj0UZsf/JvGHRypCuxscu/zgm1gZRr/XuZ3tufkdq68dq0 lHEj8MY+AxRu/xYoBTy/sBm9E5wZzUuwRvtBI4e0DWy6cRaoVbtYIPxlCmwTHMRw3+ZtSNtypeu+ 9Ua+RbKxsTpw8YvAkyElgMYlvOuz45mX23W+CcQmkbFBVOwaOQJvIadsSDxvBrTFYvK+iPrA+Ytr q+E/Ooyo9lZuOuxxEIS8pZVPDRqhJWwlCBwiEUsjTvxuJJRw/uazQ6dzU8wgLrTDOVdAQvdwZXI8 KWyCHBzGHyckDifCCNxCIEQl/hYBKDrsxz8nHkg8bGMLvQincQu8pu3bxeQGP5OMbJw8JkACbvxz cFcME4V8yQLnGkRsor6zHU2iyTxHJL/McXDycSony/EzcmtMIwbXwlQscyf0Jgf8c53swqMscxZH wlXMvQTEcfOjyihHwGL8bd7sJwd8zXqSy293dB85BEqXuMxMcgFHza67yLbRcS13HGBiy8UMydXw bytHyk9CypQMJ+acJgVtJiGS0HVske0rz/Tb0O6BQAkdzQqdslX2yDAUKOi8YVT7nvWkdDu0qyHd o3w30mtm0p6K0kmo0ofH0kvm0m3WsDDNZDNd0kcwiLPSJP230zz9Kjjd00AdsdSU9NOFQgwe/bQB 2CmRygeZh6Jq69RwOYBn2igczWOnss7SadRnB9XSoij9QtVpJtXth5DZBKuAw5ADiNYnqmNXLY6c J5e2+xRD69ZPSHfrvHaDZ3dqHUZXPaA0+2xHhNdHXQyDbUV9XSmE3XaIdg9uRFR2PZRerbZJLdiB aFIpJU1ELdSavdn+R9ac/dmgvUiZHdqkXdq94gfGl9qqvdqs3dqu/dqwHduyPdu0vXoRAAA7 ------ =_NextPart_000_01BD43A6.A33F7900 Content-Type: image/gif; name="lsnpm4_span_ch-week.gif" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="lsnpm4_span_ch-week.gif" R0lGODdh9AGHAOMAAPX19cLCwmRkZP8AAADMAAAAAABmAP8A/wAA/wAGBAQBEACoEEAHBAQAFAAL BQQBCCwAAAAA9AGHAEAE/jDISau9OOvNu/9gKI5kaZ5oqq5s676TIAF0bd94ru987//AoHBILBqP yKRyyWw6n9CoQBaI1grYa5a25XaHBWsuLC6bz7YBes1+kttu43QGF7/XhDyhrs/XAQV6fwBqg4Y3 d4eKSIlBc1WLTI1ifX5wgYJ8mXWFhpORoGifQaM9jz1kpahcV16trESqTpibeJV7bbe4bARqtTd7 lrxlu7t1sqE6yDmnPKnJyraVwGa61MS0wjTaUL19ObrG1ZR5mADi0HixclTOgO/piOfb5Lc1v08E 2cbfUeE2/bp5C7jt3zl06woC5MYkHMJ4TvTd4/fQxjIczXZcdLdlUkcw/liwlIP1DF7JkyT1YVq5 ZyXJlyZhnixH852+mihj6kR5s6dLQCN3yhxqshfLoj5p+iQ6M+lNpDmjMlX685lEoVilAhX5VKtX plxGLl26M+RGGhmjQcSx71fFHw4NCvxnL5/Sb9bsDuwXN6+Qafam7cDHo9K+iYSVxC34VrEgvolf EUlr0ezZQW0faybS1+HByED0ZO5Mz0hnuZ+P9BlwunUQw6cBLgQNrO/nug1bo2NYpCbpa5fRttNo 1ojHMe+Cj7Ydus9y16Wb66aLi/fg6Xit0WachzX2xT2+f4crnvqQW883p15/fjx6dnTWCkn/W0f5 6eT1ILivPTxs/rhd/rcagAHWRmBs9t0m2oGC2QcgdwS9dmBwAFAmnw+Y7FeegfppmAcCGnr4oH0g 7odAWyISkCKCDqpY4ocrbthiHwiw9mI4IOrGYTi05NihjPTcYqKPDBJWpGfBNFbQiQBSaKEdrrAV 5RADVImjiESWqOWN6LUW44cuwrjil/f5aGJdNJ4WYiVaqujimSsuuWV3bJZH5DlbEtkXl2C6qR+N ZuZJJqBC4vjjoHr2BGiIN+a4pnk5eJioLo46etVkw114hGXwsPGGKhSiUtwxkl3yhamlhIrhqKhm oWokT2oq66y01mrrrYvEiuuuvPbq66++6qpEIqkcB8irwCar7LLM/tYhrK3I0hrIQYYoKUYnmDUL bRJOZvprtEAcBkeDuWxnBbbaagquJ/BBolaoH7kDCxi8kCuNddjYi4Yv+CrE7LrpYtguR52uBXBh BZ6haL/+FMhwEnshFAx7khRGjB/WpntHxhYN7MypEG1MSVuyXWwvx7355a8VEYOjcj7XNIzXHwef FzMpHiPHbRchEUdhGBE2gSTFepGrr2MLU/SwaRErzRxn4tRiLtR3oSz0IZZIzXC38ek8K9CJWY2D bUHnRrKCYvuwWMK5eReQjlRLJNjRiLX38tQpB7g0EldBtnXOyoAcz3IT8V01eHgjLF5uSDo8j2p0 BobdX+PdM9ve/mhLXjbSJOubtssYPy1lEc+yanDjbCt+oOWU37cyys7hZzTHAzIonXvZ0b1j0sx9 jhh+jwvje+69f8a1u4hYVjMU9KEm4JFEq968dtXR7np9qjc9YvaHb8/96o/553rmaXeP3fHKTg/e 2IAW8OiDjS065/XiE1rk927fD/qRnu3Pv/O/i52O+me3JgHOV+47U59ghKA6xQVOgZofXVKzwAe2 aUhDIpvl0lQ1Rr1pUoXCB6V6IakONapSDCQIpbAUJj+hcE+aC2E5SvjBMUEncy/KIQhjszlgKPCD YtrhpYbwLDfw7A49K0KV0GWF5c1HEU7E2SGiGC4oWvEPRfyW/mTC8KkocDE5SexiE7d4LMEtoYto LJUXO/XFeY2xjGJcY8dGB6V5tbFga8liwPbIxz76cVd6/KMgB0nIQjrLW4ZMpCIXycgmBNI4nzJj GRtJyUpaMhSPpBcr4pU8N/5AeaAMpShHScpSmvKUqEylKlfJSlIOoJWwjKUsZ0nLWtrylpLcQSbl Q0VoZI1mj/sDEwnZyz6iL1nFbMXENEEtTfhOCcO85BmSWYZjAouavztGD032TCRE01PSJBUjDtgr bCrImZibReKc8M0/mlNb1tTiGsRVrnUuIXVm4BdcwjksBJKzY3eUYiOOE1ApogGfLMsGOu0JMX5V JEnMqxYl/l8VT6+pq17blFlGu0FPPORvMM0cBiVsws8qkg6RbMllKKiZtHEhVKNu6WZ4Pjq2c4oU psu8RDLEVlF5eHJwJcNp1KrR0WCyzGEyLQxNd5dOgbHuqXMJHyGHyNN/VkalkQBadKL6tqR+z0BH LWpTKbfUAEImIlHzKocUOlV+3EwtmOqaRmQFtrE252zRE5rKdAe5vTKUPPlzGprUqjW1Xo6vdRyE 8Cji1Lgib47RaqPgilWEuhr2sJojR0cRmze9/VVtTcMsAG9HhsIK1a6GeylHGZNNeTl2FZs6YiSv apzbHBV1lzXrYjdKNbzyNmWBFe1oQauZTPA1p6QNGmrz/nOyalQva7Tp6VWxuojOvRVqzbvtYH/b Oupxt3WBLQbcupsk424zcaL7rPh8m1uw+q0x0m1FQQfnvGdWTr2IUR9h18Y2+0buvSy6He/Kqw2n kde7+J3RdttbkGkBZpvxtaO6CIixzspuucINcGoHjBrkFjC8BL4hc8drVM4GsHOAKVoMd8O4EL/s cRHG44UIl7sC6vfFEhofaoGU19uBmMf1I/Fup0YaAiKNwydLKvE6DGOr/hQkaqSjcnBrYggRCKrE RTLZinHgAaowY5X48eKyvCD9Bfm+WBYwdKpXN+xe2DlOrpiEj0XGWIx5xOCD3ZWhm2PzURnH7NOD mO/8/rz/La1L8Kvw93j4XNbGzctwPqlcy7mh+P3HdnjWn6UNvZ1bDLpyheb0Q8n3v+/p18p2ZVCM oXXlGQ1qfEF235/4Y1T/zRrTof40qAONaOjZWtRvafVw/XPj90j6sZT2npUZlcHVFeZRknr12s4s a2mTmKk1qqCydTvCVvO6T8XWMLfNTDtr92XVtQoEn0AtwwyJydu1eZ8MWwg8l1nqR2Eyd4qxTUJ5 /wmFbZqdkAKebxOO2cUVlKC2J4fqMKlbiIw+9A/Lg+5ZSDKJsYh2Bl9I8IHrO4JseuGD8V1mh7xP 3jGCEMRzaD83pehOmH1R/jquJg/pdkWyThMLJ2Wp/levW36H0uGrUW2mgb/75djTQY6q3XJ6g+lE cTbikzn5gyUuEZdYz7rWt17LV3L962APu9jD/tqSmv3saAfWLm8VSU4uT4wYN10c5ozxOEj2I9St rE+TQ2c50haOb5dH3uUMxzlfaO3bovvgNWkS2UYRibJ9Y+NLImNJRJ6NTXR8Zfw+dTPEsVjvPALi 00760pt+DVNIvepXz/rWu/71sI+97GdP+9rb/va4z73ud8/73vv+98APvvCHT/zitx4GyE++8pfP /OY7//nQj770p0+B0Z/++tjPvuhRqv3ue//7UQe/+Mc/fuuT//zoD6f5YWv4OVY+/fCPfzLW77Nn /syX76GXv/73rwP6w/WOlNV+BsV/+ZcGjFSA6RN+0iJIb8BgF9JOosB/7Fdbx5aA86QIoSNRgwCB BOhPFYhM41JrthBSbWAOG3iA31dxX+NSGGYc36VXLXgEHOh5EuhaFFh21wROL3hPO8iDMWgEM6h/ CEh4RMR9u2JOVYZWCVZZS6hEKOh9KkhX9fSDYNCDR3aCU1SDczVOH5iD85SEMEiFBYSFWkhHR6iA K3gvYthnDjgYbAUHhdCGvJSCaMiFnDIGi+d+CgOGkmCFG7aGQaBPahNRZWiGeteFyHF/AgN6nxCA jDdNfHhkgJhpnNBjhYN+Q3hGcfZFl3GHUdZ3/mdRHF+RFVNBFqN4ijJBFWBBiqxYLKa4iqi4E2pQ FS9xFbF4i7CYizGhEq2oi7jYi8D4i6dYd0Dgf64Vd5D3MxjVhK9RVHsYiQ3VQxA1dwkyUltFTNyy iXw3Y54CjX84iWc2VMRQVtQAjhZDDtdYgjuVEDiYiO+XDCylWipmjm6YZONYNhPDjPukWSKYSFWF iB74jH5oGmLlXPZICeQYJANpUTWlXYqGjWnWWEU4ae6XiZ+EBwXJTW5hkBFCj+eQkORjjRHZYg+p jho4kjYoBMISRwYTVFwlNRy5kZp1kCwDksOmGNcFg+Ygh/Egjk8EkOXkkvNYYESVZAxWX23Y/jJM tYYNmFanxZPpQFXsOJHIBlDvuFLBMzK44ZGZA1anVRvkkJD9s4bcAI5GBk7V4lZCyZCOYISbdJWg oFX9SJKboI/fVmB2ORGblZcFEVwZtjI8OBsa9YaCtFgLIZEq6ZZsZ1tPOQ982WACN4lGtpDX4ZcN J48uAwum9ZKPGR5YU1p4CV9WhSxp1HlPJjCU6SDW1ZkdlppLaWANY5mk5o1dpTSNWS7O6FyOxphb 2I7JI3ekYH+Vx4l655p3OXJH6Vd5WXJ1+Zj/FZpJJ2C7aYkbRpgoKYmd6SC7mVFRyI4AOCUsiTPG +ZqhY46HY1ac2ZzmGVrjhplrxZvU2Vf9/lUP3qgaoAldfwOUenhR47kQowGfStgg7smGAuqc7HmZ AxpANnVccxlqIuSA57mWdsEdFJqSbUmR+zmHfJGTXXaWYTg39SkgKNafjBFe7Zmg5wSioIFeArSZ 6WmYJsOY3FmHE3Y33yg6aJUZaPOiCDcXJjqbgCY9O3pc6fQ0Ieqg5pWUT/FeE5iYGPqWFokh9ZVa 0zOd8olgdslfRwo6P8pwBIqlz9Oh6mmW3VOhGiWjEKaNigiPU3qlDfSDQFadK8Yb/qU9CBqkCkYd MNlaAjaifIk4rFlc5dkv3TljFDY89wWnfuahTPNnQ1WnIAakePpt2QVVSVhkW/prgspi/jzoYDia ocWomOFJXwYhVTYGaS04Ioi6a59VO86WadgjWF+KnJmaX9GpqIMKQLgQY1FKCrhFgmoGaxaWqI2q ZQInn7p2bYuWXuL1W5e2qca5a8BarJIqG7tKo4aqp+VVQN5mNcwJNzt2Z1/WZ8kqYrHGIknKWxGH oki6PoAJOZ5qpNTCq/3kBcSoBT/jqHr2rewmpuDKZbO6rozlY2YWjq5Bq4dGaw3apw30OvIpqb8k GlFHUfjaiFMyH4RmsLC2r7QmPN11arMjYOUqbprqa9QGPADLrWv2OqmGO5Hmm4S4d5sHlynFFSzh FLSYFVSxszjLs0CDij0btDfrErZY/opD67M4gbRFuxNK4R1HK7Q7S4pQO7VPOy1eobRV67PnYLVG i7VJCxS2GBSn6LVJAbaBAJw+ECvEUlmX13f26oLSmj3hNnIBe2HvuqwCO6kQMrIkS564hrfXU5K2 dmoh+Wj3EV9rm61xW7J/q7GqSmZzy6jfxrfR6beNm6eRW1wLK2x6y7jTUai8pLDtWrCFlrlz47iX e5eU+6lLabr4hHCGBrhvyq6L2hqgezqiO7j8M2IQNzmbJmpkpj3bdqImW1PAdl69NrvQqGrYGrrD qyC9uzi/20LRq0Hco27Fe5yrW7kQS7rj5rqAlmede5zicbshA2/L1mzV27e/wyRF/kc/3/NzwrqU 28u6pAa+dHO8qXO811kbpmu+pJq7OEQTCje/8TZxKXRr5sprG+e9N+cdQrS+u6EbSLe4HifB9da6 uFZuBALA6fBw8ztvfYLAGTsRGLRCTce+n2EiPaJzJWxWIDIAIrcoswZCS9lxXILBKoQ2HpJAPBen POxA9CNxOmxsMMsr2Esm0hZAKxco/9a7jgkQRZcZGqfA7hpvhpJv+maqN0dCVjzCQOR0gnVrTOJC /3ZwOMRCCbfFW4nCOhS4EqcoevK+uuDB0JBzlVLFHGRBLRpyDRxtW7bHeyJ0efIbRufGgtIZInLD EcNsWYJyi0x0L/fILtTEH2dy/kfnx8qrc2X8xUDsQ1pcQ2/iEHb8k8P5iRhiJQwUcIVMczT3QCTn x2occt9xcgv0c9zBJ/tAxyNMcLjMwxoSWPLLx9Aqxto2cQondBakJ7uMJa08zMu8cC73xu+2ZUpn cLHsJ0FUyjhzf8QpBFb3YxV8wTvkuo5cQ/KbuRckKIX8rGAMdBiUx+x8Qc+pzbU8z+ncbRzUzk+H z3VMy0Isy/jMz92Tw6I80AjtIwmkyHOyH9xsUo1XZ4FodRGYhRbNLhedLRkNTFikmLdSJYUY0iJt A8Y40iZ90odQ0iGDyokFKol1DFyUf6Na0e/4eBdZTSw9KyoNVBI20/W6jUBtc9M9ndObIl/yRXlG RHkxfZrDAjI+fUZhpNQ0ONRMPX8e7YWY99SMsNSJS7NFTdUtndVEfYPfqdVfHdRjfdamGdZvWdWY dNUoHddybdVPOtd2fdcpDdd4vdd8LQZzYHyAHdiCPdiEXdiGfdiIndiKvdirFwEAOw== ------ =_NextPart_000_01BD43A6.A33F7900--
Subject: Re: (usr-tc) Dual HiPer ARC's
From: Brian <signal@shreve.net>
Date: 1998-02-27 22:06:06
On Fri, 27 Feb 1998, Mike wrote: > At 03:07 PM 2/27/98 -0600, you wrote: > >Do the two ARC's have to use seperate IP pools? Or can I delcare the same > >pool on both and have them share it? > > 2 Pools Will be required. > > >Does MPP work between ARC's within the same hub? > > No. MPIP is not supported in the current ARC code. > > >Is there any concerns or caveats I need to know about using 2 ARC's in one > >hub? (that's not clearly expalined in the docs) > > You must turn off Chassis awareness. This is done by typing > 'dissable nmc chassis_awareness' at the ARC CLI. > gotcha, then I would suspect I would have to assign all card types/ownership statically....... Brian > -m > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Vendor-Specific RADIUS attributes
From: Timothy A. Gregory <systems@tarjema.com>
Date: 1998-02-27 22:13:20
We just moved to the Livingston RADIUS 2 (and the *do* have a Linux version!) and our detail logs are very complete for Total Control Hubs, Portmaster 3, Portmaster 25, Ascend Pipeline 400 & 4004s. mebbe it'll help. ftp://ftp.livingston.com/le/pub/radius On Fri, 27 Feb 1998, I. Dwayne Koonce wrote: > Hello...I'm new to the list, so bear with me if this has been asked since > the last message in the archive... > > We're currently running Ascend RADIUS on a Linux box for authentication & > accounting for two TC's, two Ascend's, and a Portmaster. We've just > decided to expand the information that we keep in our logs (in a MySQL > database, courtesy of homegrown patch), but some of the information we > want is sent by the TC's in the Vendor-Specific attribute, which Ascend > radiusd doesn't support. > > I've tried Merit AAA 3.5.6, which appears to go out of its way to support > vendor-specific attributes, but it logs them as an empty string and > outputs errors like the following: > > Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown > Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific > attribute Vendor-Specific=vUSR-000000c700000001 > Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown > Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific > attribute Vendor-Specific=vUSR-0000901900000000 > > ...which I see from the list archives that some others have as well. In > fact, it looks like there's been quite a lot of discussion along these > lines, but the threads appeared to mostly die out with no solution posted. > > So, my question is: Is anyone currently using these attributes > successfully (with source-available software)? If so, then what? > > > ____________________________________________________________________________ > I. Dwayne Koonce E-mail: dwayne@txcyber.com > ____________________________________________________________________________ > "It's dangerous to be right when the government is wrong." --Voltaire > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > Timothy A. Gregory Northwest Link Systems Administrator Arabic > English Translator
Subject: Re: (usr-tc) Dual HiPer ARC's
From: Curt Shambeau <curt@execpc.com>
Date: 1998-02-27 22:24:02
> > You must turn off Chassis awareness. This is done by typing > > 'dissable nmc chassis_awareness' at the ARC CLI. > > > > gotcha, then I would suspect I would have to assign all card > types/ownership statically....... Yes, and RE-CONFIRM all those settings twice after setting it up and saving all the settings, and rebooting the box. We assinged every other card to an ARC - all the odd numbered slots on one card, and all the even numbered cards on the other. Has been working out OK for us so far. Don't forget also, if you are using T1 lines instead of PRI's, you need to set the card type (i.e. SET CHASSIS Slot xx card_type HDM_24), but you also have to set the number of ports on each card, because it defaults to 23 ports, not 24 (i.e. SET CHASSIS SLOT xx ports 24) | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |
Subject: Re: (usr-tc) Dual HiPer ARC's
From: Brian <signal@shreve.net>
Date: 1998-02-27 23:06:32
On Fri, 27 Feb 1998, Curt Shambeau wrote: > > > You must turn off Chassis awareness. This is done by typing > > > 'dissable nmc chassis_awareness' at the ARC CLI. > > > > > > > gotcha, then I would suspect I would have to assign all card > > types/ownership statically....... > > Yes, and RE-CONFIRM all those settings twice after setting it up and > saving all the settings, and rebooting the box. > > We assinged every other card to an ARC - all the odd numbered slots on one > card, and all the even numbered cards on the other. Has been working out > OK for us so far. great way to distribute load........ > > Don't forget also, if you are using T1 lines instead of PRI's, you need to > set the card type (i.e. SET CHASSIS Slot xx card_type HDM_24), but you > also have to set the number of ports on each card, because it defaults to > 23 ports, not 24 (i.e. SET CHASSIS SLOT xx ports 24) > yes, I have learned to just give the FULL parameters for setting statically, that way I am sure it knows what I want to do. Thanks, Brian > -------------------------------------------------------------------------- > | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | > | Executive Vice President - Exec-PC, Inc. | > -------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Reported slowness
From: mt <tsaim@mft.com>
Date: 1998-02-27 23:09:15
We experienced same slowness problem. I also found that the slowness only happaned on "Download". I test ftp "GET " and "PUT" with the same file on the server. Get first , then I put it back. The Get is always 1/4 the speed of the PUT. And the PUT performance appears to be normal. In another word, the slowness only happened on the "Download " stream. The tested file is 380K. Download it took 300-350 seconds. Upload the same file took about 70-80 seconds. The modem I used is a microcom 28.8 desk porte. There was only 3 ppp users on the TCL hub when the test was performed. The netserver card has 16MB and run TCS 3.01 set of code. BTW, the FTP server I against during the test was sit in the same subnet as the TCL netserver card. Any idea ? Thanks Meng tsaim@mft.com John A Timon wrote: > Hey all, We are having an interesting problem; in that > some of our users are reporting _very_ slow performance on their > connections, I was suspicious of their modems at first until i > experienced it myself. I am using an USR 56K sportster, and on this > last connection I noticed very slow transfers :( I am also getting > periodic busy signals (although the chassis is less than 1/3 > capacity) > > any ideas?? > > tia > John Timon > The early bird gets the worm but, > the second mouse gets the cheese > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.
Subject: Re: (usr-tc) Vendor-Specific RADIUS attributes
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-28 00:13:56
I. Dwayne Koonce was heard to say: ... >Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown >Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific >attribute Vendor-Specific=vUSR-000000c700000001 >Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown >Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific >attribute Vendor-Specific=vUSR-0000901900000000 > >...which I see from the list archives that some others have as well. In >fact, it looks like there's been quite a lot of discussion along these >lines, but the threads appeared to mostly die out with no solution posted. > >So, my question is: Is anyone currently using these attributes >successfully (with source-available software)? If so, then what? I recently "fixed" 3.5.6 to deal with the way USR does its VSA's for a customer (big ugly mess, but ...) I'll post the rather trivial patch here when I get home (where that patch currently lives.) --Ricky
Subject: Re: (usr-tc) Vendor-Specific RADIUS attributes
From: Ricky Beam <jfbeam@interpath.net>
Date: 1998-02-28 00:13:56
I. Dwayne Koonce was heard to say: ... >Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown >Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific >attribute Vendor-Specific=vUSR-000000c700000001 >Fri Feb 27 14:01:26 1998: generate26: Vendor 429 attribute 0 unknown >Fri Feb 27 14:01:26 1998: gen_valpairs: non-encapsulated vendor specific >attribute Vendor-Specific=vUSR-0000901900000000 > >...which I see from the list archives that some others have as well. In >fact, it looks like there's been quite a lot of discussion along these >lines, but the threads appeared to mostly die out with no solution posted. > >So, my question is: Is anyone currently using these attributes >successfully (with source-available software)? If so, then what? I recently "fixed" 3.5.6 to deal with the way USR does its VSA's for a customer (big ugly mess, but ...) I'll post the rather trivial patch here when I get home (where that patch currently lives.) --Ricky
Subject: Re: (usr-tc) opinion needed - mass software upgrade.
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-28 01:41:00
At 12:07 PM 2/27/98 -0700, Pete Ashdown wrote: >Allen Marsalis said once upon a time: >> >>At 10:34 AM 2/26/98 -0500, Mark R. Lindsey wrote: >>[snip] >>>(I'll admit that I'm not familiar with what can be done with tcmunix >>>insofar as scripting is concerned; I've yet to find the Linux release. >>><ahem>) >>> >> >>I'd be happy with a release that works with Solaris 2.6.. > >I was going to ask this question earlier, but I presume the latest TCM does >not work with 2.6? > That's the way i understand it. We purchased a sparc5 just to host the unix tcm and it came with 2.6. It's been a couple of months but brian mentioned to me back then that tcm worked on 2.5 and he refused to downgrade and wanted to wait this one out.. AFAIK we are still waiting... allen
Subject: Re: (usr-tc) Reported slowness
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-28 02:09:36
At 02:16 PM 2/27/98 +0000, John A Timon wrote: >Hey all, We are having an interesting problem; in that >some of our users are reporting _very_ slow performance on their >connections, I was suspicious of their modems at first until i >experienced it myself. I am using an USR 56K sportster, and on this >last connection I noticed very slow transfers :( I am also getting >periodic busy signals (although the chassis is less than 1/3 >capacity) > >any ideas?? > This happened to us around six months ago.. even power cycling the chassis would not unstick it. We eventually reflashed all code, configured it *exactly* the same way (from scratch), saved to nvram, and reset.. Never happened again.. It was the exact same code revision and exact same configuration.. We thought it was telco related at the time because of the busies.. It was the only time I remember crying 'wolf' to the telco when it was on our end... but they never forget... allen
Subject: Re: (usr-tc) ATS0=1 hangs up immediately
From: Tatai SV Krishnan <tkrishna@bubba.ae.usr.com>
Date: 1998-02-28 07:42:06
On Sun, 1 Mar 1998, FBSD wrote: > > We use a NETServer 16/I for dial-up access. I would appreciate any > pointers on the following: > > 1. Right out of the box, there is a default init string called > "USR_int" with the command of "ATS0=0" However, the modem(s) ANSWERS > incoming calls as if S0 is set to a non-zero value. I've checked by > issuing an "ati4" and confirmed that S0=0. Why is it answering when > S0=0? There is a reason why thee is a INIT script in the first place. The init script has to set the modem in a proper way for it to answer. Here in this case, when a call comes in, the modem senses the call and informs the above layer about the call - which will then tell the modem to answer. > > 2. If I have the modems use an init script with the command of > "ATS0=1", the modem will answer but will IMMEDIATELY hang up! Is this > a known problem with NETServer? "ATS0=2" seems work fine. > ats0=1 is to tell the modem to answer on one ring, which is good for a modem connected to a pc, not this this case. The modem should inform its above layer about the call, when you set ats0=2 it will work since the modem sense the first ring and the calls is told to answer from above krish \ T.S.V. Krishnan \ \ Network System Engineer \ ( : - : ) \ 3Com ............ \ ----------------------------------------------/ tkrishna@bubba.ae.usr.com ----------------------------/ http://interproc.ae.usr.com ----/ Any Sufficiently advanced bug is indistinguishable for a feature. - Rick Kulawiec > Any ideas? Many Thanx. > > fbsd > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >
Subject: Re: (usr-tc) Routing?
From: Bob Purdon <bobp@southcom.com.au>
Date: 1998-02-28 09:13:07
> Yowzers. Doesn't look like there is any way to create "perfect" routing > with your setup (at least not without money/upgrades). That was my opinion too :-( > 1. On your border router run a decent interior routing protocol, such as > OSPF or EIGRP. (I'd suggest OSPF myself, better compatiblity with future > equipment) Yeah, that's the way I decided to go, and run OSPF on the Unix boxes as well. > In summary, you end up routing a *very* few extra packets through your > border router. It's not perfectly optimal, but the performance loss is > so minimal, there really isn't a point to worrying about it. Yep, and that's the way I decided to go. I've got another couple of Unix boxes to convert over to OSPF and then I'm pretty much done. Thanks for your feedback - muchly appreciated. Cheers, Bob.
Subject: (usr-tc) Scotty/Tkined
From: Timothy A. Gregory <systems@tarjema.com>
Date: 1998-02-28 10:04:48
Has anyone used tkined to monitor/work on total control chassis? If so, would you be willing to share info on how you did it? I'd just *love* to rip Windows off my PC, but so far total control manager is the only way I have to do all the funky stuff to the chassis that I need to do, and they don't have a version for FreeBSD (or Linux/SCO/BSDi). Timothy A. Gregory Northwest Link Systems Administrator Arabic > English Translator
Subject: Re: (usr-tc) opinion needed - mass software upgrade.
From: Brian <signal@shreve.net>
Date: 1998-02-28 11:42:49
On Sat, 28 Feb 1998, Allen Marsalis wrote: > At 12:07 PM 2/27/98 -0700, Pete Ashdown wrote: > >Allen Marsalis said once upon a time: > >> > >>At 10:34 AM 2/26/98 -0500, Mark R. Lindsey wrote: > >>[snip] > >>>(I'll admit that I'm not familiar with what can be done with tcmunix > >>>insofar as scripting is concerned; I've yet to find the Linux release. > >>><ahem>) > >>> > >> > >>I'd be happy with a release that works with Solaris 2.6.. > > > >I was going to ask this question earlier, but I presume the latest TCM does > >not work with 2.6? > > What gets me, is compiling TCM for 2.6 should be trivial, and lead to relativly few issues which can be immediatly fixed. How old is 2.6 now? Sun is about due out for a new version............ > > That's the way i understand it. We purchased a sparc5 just to host > the unix tcm and it came with 2.6. It's been a couple of months but > brian mentioned to me back then that tcm worked on 2.5 and he refused > to downgrade and wanted to wait this one out.. AFAIK we are still > waiting... > > allen > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > /-------------------------- signal@shreve.net -----------------------------\ | Brian Feeny | USR TC Hubs | ShreveNet Inc. (318)222-2638 | | Network Administrator | Perl, Linux | Web hosting, online stores, | | ShreveNet Inc. | USR Pilot | Dial-Up 14.4-56k, ISDN & LANs | | 89 CRX DX w/MPFI, lots of |-=*:Quake:*=-| http://www.shreve.net/ | | mods/Homepage coming soon |LordSignal/SN| Quake server: 208.206.76.47 | \-------------------------- 318-222-2638 x109 -----------------------------/
Subject: Re: (usr-tc) Reported slowness
From: Robert J. Adams <radams@siscom.net>
Date: 1998-02-28 16:28:04
John, Did you find the fix for this.. we are having 1/2 our customer base scream the same thing.. we are running the most current code as of 2/2/98. -Jason -- Robert J. Adams radams@siscom.net http://www.siscom.net SISCOM Network Administration - President, SISCOM Inc. Phone: 888-4-SISCOM 937-222-8150 FAX: 937-222-8153
Subject: (usr-tc) QuakeWorld Lag info...
From: Jason Brunette <root@tcbi.com>
Date: 1998-02-28 17:29:13
Having read through the usr-tc archive, I've noticed that many other people are having trouble with their TCs and Quake performance. I did quite a few tests with QuakeWorld, and have come to this conclusion: The more packets being sent and received by the user playing QuakeWorld on a TC, the worse the lag is. It sounds kind of obvious, but let me explain... 1). In QuakeWorld, there's a client-side command called "rate". This command tells the QuakeWorld server how many packets to send to the client (although I'm not sure what the measurement is). GameSpy configures the QuakeWorld client with a rate of "2500". This is the recommended setting for modem users. Adjusting the rate command can make game play smoother by setting it higher, if you have the supporting bandwidth, or can reduce client-side packet loss by setting it lower. This "rate" command is one part of the TC Quake Lag equation. 2). The "rate" command isn't the only one that determines how much data is sent and received by the QuakeWorld client. Performance of the QuakeWorld client is also a key factor. The higher the frame rate (or fps) of the QuakeWorld client, the more data is sent and received by the client. The lower the frame rate, the less data is sent and received. Knowing these two things, I went to town... I did tests on a TC with 20 PPP modem users (no ISDN users) with an X2 connection of 42k/24k connection, and more tests on a Livingston 2e with 29 PPP modem users (no ISDN users) with a 26.4k/26.4k connection. The TC is 1 hop away from the QuakeWorld server, the Livingston is 3 hops away. There is no bandwidth difference between the TC and the Livingston in respect to the connection to the QuakeWorld server. Both are via T-1 (although the livingston is farther away). The QuakeWorld server is stock 2.10 server running no mods and is on DM4 with no one else but me in both sets of tests. The QuakeWorld client is running on a P150 with 24MB mem and a decent video card (non-gl) which gets 32fps with "timedemo demo2" at 320x200. I set "r_netgraph 1" to display the lag graph. The TC is running the latest TCS 3.0.2 code, Quad modems, 16MB netserver, dual PRI card and is connected to the network via a Cisco 2501. The Livingston is running ComOS 2.7.2, 1MB mem, MultiTech ZDX 28.8k modems and is connected to the network via a Livingston IRX 114. TESTING THE TC =============== I started with a rate of 2500 and ran around the map. Pings were quite good for a 42k/24k connection...hovering around 155 no matter where I went or what I did. 1% packet loss throughout. Very good. I changed the rate to 4000. Started noticing extreme lag depending on where I was in the map. Wide open areas where much of the map could be seen had great ping times. However, enclosed areas (like the teleporter at lava-level with the health and 100 armor) display huge lag (800-999) but a very sequenced, patterned set of delta errors (blue lines on the lag-o-meter). About every second the delta errors would come, last for a second then go a way...a second later they would come back, last for a second, etc. Even if I walked into a wall and stared right at it, this happened. This was 100% reproducable. Packet numbers where high according to the QuakeWorld server (shown by typing "status" in the QuakeWorld server console and observing "packets/frame"). When staring at a wall with rate set to 4000, packets/frame averaged 0.44. If I turned around and faced an open area of the map, the lag went away completely and packets/frame were down to 0.32. With rate still at 4000 and standing right next to a wall and staring at it, I changed the resolution of the game from 320x200 to 640x480. Frame rate went down quite a bit (probably 15fps or so). At this resolution, staring at the wall with rate at 4000 had great ping times and no packet loss. No matter where I went again, there was no lag at this resolution. QuakeWorld server reports packets/frame at 0.20. The problem is quite clear at this point. I felt the results of the TC test were complete. TESTING THE LIV. 2E =============== I again started with a rate of 2500 and ran around the map. Ping times where respectable for the connection speed, hovering around 200 or so. No matter where I went or what I did, ping times didn't change a bit. Same as the TC. Changed the rate to 4000. Ran around the map. STILL no lag at all. No matter what I did...the ping was still around 200. The QuakeWorld server reported packets/frame at an average of 0.50 (higher than the test with the TC). So, connecting to the Livingston 2e resulted in a higher packets-per-second count, yet no difference in ping time. Changed rate to 10000. Packets/frame fluctuated from 0.50 to 0.60. No change in ping time. Still great. I expected the higher average ping time of 200, but I definately did not expect the Livingston 2e to outperform the TC in QuakeWorld performance. Conclusion =============== The TC chokes at high packet rates. Recommendations =============== For QuakeWorld users, set "rate 2500" or lower. Additional Info =============== During the lag on the TC, the following debug information from the netserver ("set debug 0x51") is displayed constantly during the lag (we never get this output at any other time): usr2> set console Setting CONSOLE to admin session usr2> PPP recv: BAD CRC Ptrace: 21 45 00 00 37 93 49 00 00 80 11 0D A3 9C 2E BE 7D 9C 2E A2 EF 69 79 6B 6C 00 23 77 A6 78 07 00 00 52 07 00 80 03 81 B2 F6 51 C9 14 81 B2 F6 51 C9 14 81 B2 F6 51 C9 14 PPP recv: BAD CRC Ptrace: 21 45 00 00 37 97 49 00 00 80 11 09 A3 9C 2E BE 7D 9C 2E A2 EF 69 79 6B 6C 00 23 72 A6 7C 07 00 00 54 07 00 80 03 81 B2 F6 51 C9 14 81 B2 F6 51 C9 14 81 B2 F6 51 C9 13 PPP recv: BAD CRC Ptrace: 21 45 00 00 37 9E 49 00 00 80 11 02 A3 9C 2E BE 7D 9C 2E A2 EF 69 79 6B 6C 00 23 64 A6 83 07 00 00 5B 07 00 80 03 81 B2 F6 51 C9 14 81 B2 F6 51 C9 14 81 B2 F6 51 ...etc... I don't know if this'll actually help anybody, but it seems pretty cut-and-dried. High packet numbers on a TC are bad for QuakeWorld. I'd be great if someone else could try this as well...since it could be just me. Jason Brunette - Technical Support Support - support@tcbi.com TCB Internet - http://www.tcbi.com/ Personal - jbrunett@tcbi.com 920-451-7776 - Fax: 920-457-6616 Serving Sheboygan and Manitowoc, WI
Subject: Re: (usr-tc) Scotty/Tkined
From: Allen Marsalis <am@shreve.net>
Date: 1998-02-28 18:48:55
At 10:04 AM 2/28/98 -0800, Timothy A. Gregory wrote: >Has anyone used tkined to monitor/work on total control chassis? If so, >would you be willing to share info on how you did it? > >I'd just *love* to rip Windows off my PC, but so far total control manager >is the only way I have to do all the funky stuff to the chassis that I >need to do, and they don't have a version for FreeBSD (or Linux/SCO/BSDi). > (or Solaris 2.6?) Seems like most isp's are structured more within unix than Windoz for obvious reasons. I guess marketers believe that windoz users always need canned apps and unix users always write their own code or at least support themselves.. Their product lines seem to refect this. I think venders should supply some code for unices, unsupported perhaps but with source code. I would like to see anything for a commonly used unix like Linux, FreeBSD, or Solaris 2.6 or even x86... What are the choices for Unix TCM? Solaris 2.5 and IRIX? allen
Subject: (usr-tc) Re: ANI number logging in TC and S/A server 5.0.7
From: Rasmus Helmich <rh@webpartner.dk>
Date: 1998-02-28 19:40:53
Hi all How do we set up the TC and/or the S/A 5.0.7 so it loges the phonenumber of the incoming call? All other logging is ok best Regards Rasmus Helmich
« January 1998March 1998 »
USR Total Control Mailing List Archive · Messages from 1995–2001 · Generated from archived data