January 1998

On Wed, 31 Dec 1997, Blake Fithen wrote: > Should that file be locked? I can get all the other files but not that > one. > > Blake Fithen > That file was always locked when I tried to get it, even though everything else was fine. FTP to totalservice.usr.com and login with your totalservice name and password and get the file through there. That was the only way I could get it in the past, and I just checked and it still works. > > -----Original Message----- > > From: Tatai SV Krishnan [SMTP:tkrishna@bubba.ae.usr.com] > > Sent: Wednesday, December 31, 1997 2:16 AM > > To: usr-tc@lists.xmission.com > > Subject: Re: (usr-tc) NetServer Configuration Tool > > > > This file is in totalservice. the file name is nsmgr342.zip. Logon > > to > > the totalservice.usr.com and go to software library and do a search > > for > > Netserver manager - you will get the file there. > > > > 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 Wed, 31 Dec 1997, Timothy A. Deem wrote: > > > > > > > > All, > > > > > > I'm trying to locate the netserver manager software that used to > > be > > > available separate form the TCM package. Does anyone happen to know > > > > > where it is on the USR site (I've searched totalservice) or have a > > copy > > > it they could email me (or have it online for download somewhere). > > > > > > Any help is appreciated.... > > > > > > Thanks, > > > Timothy > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the 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. > :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::

On Fri, 2 Jan 1998, Brian Elfert wrote: > Are there any known problems with Total Control chassis running in a 90 > degree environment? > > I don't have any AC right now, and our building's ventilation system > shuts off between 7 PM and 6 AM or so. > > I'm wondering if high temps affect the operation of the modems at all? > > I got a bunch of reports yesterday and very early this morning about > static when dialing into our modems. I can't replicate the problems at > all now. The ventilation system was off all day yesterday, and the > temperature got up to 90 degrees. > > (I will be moving the equipment to a cooled environment soon.) We used to run 3 TC chassis in a closest in an airplane hanger in and during summer the temperatures stayed around 90-125 degrees with ventilation and fans, so they will work, but you should still keep a good air flow or the temperature can start to affect the modems. Are you keeping track of air temperature or are you looking in TCM at the temperature reading of the chassis? :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::

On Wed, 31 Dec 1997, Tatai SV Krishnan wrote: > On Wed, 31 Dec 1997, Terry Kennedy wrote: > > > As we move from Annex 4000 and mulitech modems to USRTC's > > a problem has crept into our network. The netbios broadcst's from > > our customers are increasing on the network. the annex 4000 allowed > > not to load the broadcast and I beleive this keeps netbios broadcast's > > off the local nets. At least tcpdump shows no broadcasted Netbios packets. > > My question is, is there some way to do the same with the USR equipment. > > USR TC with NETServer does not do any Netbios - Also as a router the > NETServer does not allow any broadcast message from Network to dial-up > and vice versa. Your customer equipment can send IP encapsulated netbios > packet - This packet can come in UDP and TCP - UDP is broadcast and this > packet will not be sent across - TCP however will be sent. > > Check your Network - Use a sniffer to see who is sending the netbios packet. > > krish Those netbios packets are on UDP and TCP ports 137, 138, and 139. You can filter them, but your customers won't be able to use Samba. #netbios-ns 137/tcp #NETBIOS Name Service #netbios-ns 137/udp #NETBIOS Name Service #netbios-dgm 138/tcp #NETBIOS Datagram Service #netbios-dgm 138/udp #NETBIOS Datagram Service #netbios-ssn 139/tcp #NETBIOS Session Service #netbios-ssn 139/udp #NETBIOS Session Service Enter this on the Netserver command line to filter these: add filter netbios.in set filter netbios.in 1 deny tcp src eq 137 set filter netbios.in 2 deny udp src eq 137 set filter netbios.in 3 deny tcp src eq 138 set filter netbios.in 4 deny udp src eq 138 set filter netbios.in 5 deny tcp src eq 139 set filter netbios.in 6 deny udp src eq 139 set filter netbios.in 7 permit save all (note: "save filters" doesn't work like the manual suggests.) OR if you wanted to specifically prevent those packets from being broadcast on your ethernet you could make the rules more specific, for example: set filter netbios.in 1 deny 0.0.0.0/0 your.bcast.IP.addr/32 tcp src eq 137 -CH

> > (note: "save filters" doesn't work like the manual suggests.) > > OR if you wanted to specifically prevent those packets from being > broadcast on your ethernet you could make the rules more specific, for > example: > > set filter netbios.in 1 deny 0.0.0.0/0 your.bcast.IP.addr/32 tcp src eq > 137 > > -CH Thinking about this, what would you EVER want to come from your dialups to your broadcast? If your dialup customers aren't suppose to be running any kind of dynamic routing protocols, what is left that you could possibly want to goto the broadcast address? Wouldn't it be a good idea to just block everything destined for the broadcast address? 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 -----------------------------/

I have a quick question. I read alot about how in alot of texts, they will say, if you took a Class C address such as 208.206.76.0, and subnetted it using 255.255.255.240, that you would end up with *14* usable subnets, losing the lower and upper subnet. Along this same convention, a subnet mask of 255.255.255.128, yeilds *0* available subnets, since it only leaves the "upper" and "lower" parts. Was this just a policy for "classfull" internet domain routing? Am I correct in thinking that under CIDR these limitations no longer exist, and that using say: 208.206.76.0-208.206.76.15 as a subnet (255.255.255.240) is perfectly legal? I subnet a /24 into 16 /28's on the TC, and am able to use all 16.........whether or not that is politically correct I don't know. I would say yes, but some of the stuff I have been reading leads me to beleive otherwise and I think it just has to do with the CIDR or non-CIDR conventions. 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 -----------------------------/

Do any of you apply default filters to all of your users? We apply one to our "guest" account which disables email, pop, telnet, and ftp. Is there anything "protocol" policies you would want to deny in a default filter for all users? 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 -----------------------------/

Are there any known problems with Total Control chassis running in a 90 degree environment? I don't have any AC right now, and our building's ventilation system shuts off between 7 PM and 6 AM or so. I'm wondering if high temps affect the operation of the modems at all? I got a bunch of reports yesterday and very early this morning about static when dialing into our modems. I can't replicate the problems at all now. The ventilation system was off all day yesterday, and the temperature got up to 90 degrees. (I will be moving the equipment to a cooled environment soon.) Brian

On Fri, 2 Jan 1998, Terry Kennedy wrote: > Krish, > > You have stated that the USRTC won't pass a brocasted packet. > If they don't then what am I seeing. We running tcpdump on one of our unix > servers that is connected to the network through a switch. The only > packets we should see at that port are the packets destined for that > MAC address. I keep seeing tons of netbios-ns packets. They come > from Usrtc's or more accurately the users connected to them. They > can't know about server so I assume they are brodcasted. > What am I not understanding here? > The NETServer does not send any broadcast packet from the dialup client to network and vice versa. If you are seeing netbios-ns packets on your network, they can be tcp or udp. The NETServer does not send udp, if you see tcp then it is your wins server sending netbios name service - which is not broadcast. Put a sniffer on the network and send me the capture. krish > > We are allowing users to map drives thorugh samba or I would turn > all this off and apply the filters suggested in the thread earlier. We may > yet do this. I'll keep plugging away at this one. > > On a different subject altogether, how does one go about writing the > config of a netserver out to a file? On our annex's we simply did a > write "filename" and then a read "filename" and made a few port specific > changes. Until recently I had 1 TC rack and so this wasn't an issue. Now > that we have 11 setting each one up is kinda hassle. I understand that > I can write a text file but should'nt the TC write for me? > > 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. >

Thus spake Brian >I read alot about how in alot of texts, they will say, if you took a Class >C address such as 208.206.76.0, and subnetted it using 255.255.255.240, >that you would end up with *14* usable subnets, losing the lower and upper >subnet. Along this same convention, a subnet mask of 255.255.255.128, >yeilds *0* available subnets, since it only leaves the "upper" and "lower" >parts. >Was this just a policy for "classfull" internet domain routing? Am I >correct in thinking that under CIDR these limitations no longer exist, and >that using say: 208.206.76.0-208.206.76.15 as a subnet (255.255.255.240) >is perfectly legal? >I subnet a /24 into 16 /28's on the TC, and am able to use all >16.........whether or not that is politically correct I don't know. I >would say yes, but some of the stuff I have been reading leads me to >beleive otherwise and I think it just has to do with the CIDR or non-CIDR >conventions. That is a change that was made in moving from classful to CIDR addressing. Zero and ones subnets are perfectly legal under CIDR. You probably will want to be careful implementing it as there is some equipment out there that still won't handle it. Cisco's require the "ip subnet zero" global command to do that, and last I heard, Novell Network couldn't handle zero or one subnets, but that's been a while, that might have changed in later versions.... -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

> Are there any known problems with Total Control chassis running in a 90 > degree environment? > > I don't have any AC right now, and our building's ventilation system > shuts off between 7 PM and 6 AM or so. > > I'm wondering if high temps affect the operation of the modems at all? > > I got a bunch of reports yesterday and very early this morning about > static when dialing into our modems. I can't replicate the problems at > all now. The ventilation system was off all day yesterday, and the > temperature got up to 90 degrees. I had a set of 5 chassis in a room that got up to, and stayed at around 110 degrees. It was a closed in closet in the basement of a building. They ran like that just fine. I think, when the room got up to 125 degrees is when they started acting wierd. Netserver lockups and the like. In any case, the room is now vented properly - no damage was done. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Exec-PC Vice President, Information Services |

On Wed, 31 Dec 1997, WCLynx SysOp wrote: > With the TC hub there is a way to do it with minimal perceptable downtime to > the users. The secret is the use of the "soft" busy on the modem's DS0. > This will allow the current connections to remain active on the busied > modems, and once they disconnect, they will go imediately into a busy > state allowing you to perform the required configurations. By doing a > small group at a time, you can work through the entire pool without > dropping users. It is more time consuming for the tech, but the clients > are not affected keeping them happy. Does anybody know how to "busy out" DS0s on a PRI line? I would even be happy with a busy out for the entire PRI. The current way I accomplish this is to change the "framing type" on the span. This takes the span down due to the wrong frame type and callers roll over to next span no problem. And the span comes right back up as soon as the correct frame type is put back in. The problem with this is that it gives them alarms at the switch and some times they will take my "D" channel down to silence the alarms. Then it takes an act on God to get them to turn the "D" channel back up, changing a simple maintenance into a half day outage. I've tried doing this other ways like taking DS0s "local out of service" but every other way causes fast busies instead of rolling customers over. Thanks, Randy Doran CyberGate Network Operations

Thus spake Randy Doran >On Wed, 31 Dec 1997, WCLynx SysOp wrote: >> With the TC hub there is a way to do it with minimal perceptable downtime to >> the users. The secret is the use of the "soft" busy on the modem's DS0. >> This will allow the current connections to remain active on the busied >> modems, and once they disconnect, they will go imediately into a busy >> state allowing you to perform the required configurations. By doing a >> small group at a time, you can work through the entire pool without >> dropping users. It is more time consuming for the tech, but the clients >> are not affected keeping them happy. >Does anybody know how to "busy out" DS0s on a PRI line? I would even be >happy with a busy out for the entire PRI. The current way I accomplish >this is to change the "framing type" on the span. This takes the span >down due to the wrong frame type and callers roll over to next span no >problem. And the span comes right back up as soon as the correct frame >type is put back in. The problem with this is that it gives them alarms >at the switch and some times they will take my "D" channel down to silence >the alarms. Then it takes an act on God to get them to turn the "D" >channel back up, changing a simple maintenance into a half day outage. >I've tried doing this other ways like taking DS0s "local out of service" >but every other way causes fast busies instead of rolling customers over. How you do it depends on the type of protocol you're using on the PRI. If you're using NI2, you're SOL. We switched to Custom 5ESS for just this reason. With 5ESS, you can set the PRI (either the whole thing, or all the individual DS0's) to Local Out of Service, and it will do a soft busy out very nicely. Not sure if DMS100 will do this or not. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

If it is a remote chassis you can't powercycle it. I've that a reboot of the NetServer is usually required after changing the statslot. I think this causes the NetServer to rediscover its modems with the new settings. This is now part of my normal procedure for changing the startslot and it has worked fine for me on about 75 remote NetServers. (also a reset of all the modem cards at the same time helps.) It only adds a minute or two to the operation and everything gets reset. Randy Doran CyberGate Network Operations On Tue, 30 Dec 1997, Tatai SV Krishnan wrote: > On Wed, 31 Dec 1997, Garry Shtern wrote: > > > > > > > On Tue, 30 Dec 1997, Jaye Mathisen wrote: > > > > > > > > > > > I have a chassis that has me stumped. It's been working just beautifully. > > > No problems. Flawless. > > > > > > I do a show all, everything is A/R/P for modems that should be. > > > > > > *all* I did was "set modem startslot 2" to use S1-S48, "save all", > > > "reset all". > > > > > > And now, for the life of me, I cannot 4 of the modems to come out of > > > A/I/P. > Power off the chassis and Power it on back. It looks like there is a > contention for the Packet bus master between the NMC and the NETServer > and chassis awareness is causing this problem. So Just power off the > chassis physically - Count 5 and repower it back. > > krish > > > > > > > I had a similar problem, had to get a new netserver. > > > > > > > > > > I have saved, restored, restored from defaults, restored from NVRAM, > > > restored from factory templates, re-flashed, recycled power, flashed > > > again, cursed, sworn, sprinkled with blood, and it's still not > > > working. > > > > > > Here's a cut: > > > > > > S6 A R P on 0.0.0.0 Login/Ne IDLE 17134 120804 > > > 0 > > > S7 A R P on d01a8088.dip.cds Netwrk ESTABLISHED 6914 38192 > > > 0 > > > S8 A R P on 0.0.0.0 Login/Ne IDLE 0 0 > > > 0 > > > S9 A I P on 0.0.0.0 Login/Ne IDLE 0 0 > > > 0 > > > S10 A I P on 0.0.0.0 Login/Ne IDLE 0 0 > > > 0 > > > S11 A I P on 0.0.0.0 Login/Ne IDLE 0 0 > > > 0 > > > S12 A I P on 0.0.0.0 Login/Ne IDLE 0 0 > > > 0 > > > S13 A R P on 0.0.0.0 Login/Ne IDLE 724 1515 > > > 0 > > > S14 A R P on d01a808b.dip.cds Netwrk ESTABLISHED 747 2191 > > > 0 > > > > > > > > > > > > I am completely and utterly stumped. Copying configs from a working rack > > > to this one doesn't work. Replacing cards doesn't work. > > > > > > Any help appreciated. > > > > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the 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. > >

Brian said once upon a time: >I read alot about how in alot of texts, they will say, if you took a Class >C address such as 208.206.76.0, and subnetted it using 255.255.255.240, >that you would end up with *14* usable subnets, losing the lower and upper >subnet. Along this same convention, a subnet mask of 255.255.255.128, >yeilds *0* available subnets, since it only leaves the "upper" and "lower" >parts. This is no longer true, and I'm not sure what the reason was for it in the first place. You just need to set "ip subnet-zero" on your Cisco to route these subnets. >I subnet a /24 into 16 /28's on the TC, and am able to use all >16.........whether or not that is politically correct I don't know. I >would say yes, but some of the stuff I have been reading leads me to >beleive otherwise and I think it just has to do with the CIDR or non-CIDR >conventions. There is an RFC which allows this. What number it is, I have no clue.

What is AIP? Just wondering. Tom > -----Original Message----- > From: Randy Doran [SMTP:rtdoran@gate.net] > Sent: Friday, January 02, 1998 11:10 AM > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) Cannot get some quad's out of A/I/P in show > all > > If it is a remote chassis you can't powercycle it. I've that a reboot > of > the NetServer is usually required after changing the statslot. I > think > this causes the NetServer to rediscover its modems with the new > settings. > This is now part of my normal procedure for changing the startslot and > it > has worked fine for me on about 75 remote NetServers. (also a reset of > all the modem cards at the same time helps.) It only adds a minute or > two > to the operation and everything gets reset. > > Randy Doran > CyberGate Network Operations > > On Tue, 30 Dec 1997, Tatai SV Krishnan wrote: > > > On Wed, 31 Dec 1997, Garry Shtern wrote: > > > > > > > > > > > On Tue, 30 Dec 1997, Jaye Mathisen wrote: > > > > > > > > > > > > > > > I have a chassis that has me stumped. It's been working just > beautifully. > > > > No problems. Flawless. > > > > > > > > I do a show all, everything is A/R/P for modems that should be. > > > > > > > > *all* I did was "set modem startslot 2" to use S1-S48, "save > all", > > > > "reset all". > > > > > > > > And now, for the life of me, I cannot 4 of the modems to come > out of > > > > A/I/P. > > Power off the chassis and Power it on back. It looks like there is > a > > contention for the Packet bus master between the NMC and the > NETServer > > and chassis awareness is causing this problem. So Just power off > the > > chassis physically - Count 5 and repower it back. > > > > krish > > > > > > > > > > > > I had a similar problem, had to get a new netserver. > > > > > > > > > > > > > > I have saved, restored, restored from defaults, restored from > NVRAM, > > > > restored from factory templates, re-flashed, recycled power, > flashed > > > > again, cursed, sworn, sprinkled with blood, and it's still not > > > > working. > > > > > > > > Here's a cut: > > > > > > > > S6 A R P on 0.0.0.0 Login/Ne IDLE 17134 > 120804 > > > > 0 > > > > S7 A R P on d01a8088.dip.cds Netwrk ESTABLISHED 6914 > 38192 > > > > 0 > > > > S8 A R P on 0.0.0.0 Login/Ne IDLE 0 > 0 > > > > 0 > > > > S9 A I P on 0.0.0.0 Login/Ne IDLE 0 > 0 > > > > 0 > > > > S10 A I P on 0.0.0.0 Login/Ne IDLE 0 > 0 > > > > 0 > > > > S11 A I P on 0.0.0.0 Login/Ne IDLE 0 > 0 > > > > 0 > > > > S12 A I P on 0.0.0.0 Login/Ne IDLE 0 > 0 > > > > 0 > > > > S13 A R P on 0.0.0.0 Login/Ne IDLE 724 > 1515 > > > > 0 > > > > S14 A R P on d01a808b.dip.cds Netwrk ESTABLISHED 747 > 2191 > > > > 0 > > > > > > > > > > > > > > > > I am completely and utterly stumped. Copying configs from a > working rack > > > > to this one doesn't work. Replacing cards doesn't work. > > > > > > > > Any help appreciated. > > > > > > > > > > > > - > > > > To unsubscribe to usr-tc, send an email to > "majordomo@xmission.com" > > > > with "unsubscribe usr-tc" in the body of the 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. > > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.

Krish, You have stated that the USRTC won't pass a brocasted packet. If they don't then what am I seeing. We running tcpdump on one of our unix servers that is connected to the network through a switch. The only packets we should see at that port are the packets destined for that MAC address. I keep seeing tons of netbios-ns packets. They come from Usrtc's or more accurately the users connected to them. They can't know about server so I assume they are brodcasted. What am I not understanding here? We are allowing users to map drives thorugh samba or I would turn all this off and apply the filters suggested in the thread earlier. We may yet do this. I'll keep plugging away at this one. On a different subject altogether, how does one go about writing the config of a netserver out to a file? On our annex's we simply did a write "filename" and then a read "filename" and made a few port specific changes. Until recently I had 1 TC rack and so this wasn't an issue. Now that we have 11 setting each one up is kinda hassle. I understand that I can write a text file but should'nt the TC write for me? Thanks

Thus spake Pete Ashdown >Brian said once upon a time: >>I subnet a /24 into 16 /28's on the TC, and am able to use all >>16.........whether or not that is politically correct I don't know. I >>would say yes, but some of the stuff I have been reading leads me to >>beleive otherwise and I think it just has to do with the CIDR or non-CIDR >>conventions. >There is an RFC which allows this. What number it is, I have no clue. Don't have time to really look through and find this, but look at: 1517, 1518, 1519 and 1520 for the information you're looking for... available at: ftp.internic.net/rfc -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

At 09:45 AM 1/2/98 -0500, you wrote: >That is a change that was made in moving from classful to CIDR >addressing. Zero and ones subnets are perfectly legal under CIDR. You >probably will want to be careful implementing it as there is some >equipment out there that still won't handle it. Cisco's require the "ip >subnet zero" global command to do that, and last I heard, Novell Network >couldn't handle zero or one subnets, but that's been a while, that might >have changed in later versions.... It's also a good idea to use the "ip classless" command. This makes the Cisco's use CIDR addressing. James Nugen --- James Nugen InfiNet (a service of Infinite Systems, Ltd.) 800/848-INET

On Fri, Jan 02, 1998 at 09:50:37AM -0600, Curt Shambeau wrote: |> Are there any known problems with Total Control chassis running in a 90 |> degree environment? 30'c? no. before we got a second AC, ours used to run in 35'c (er ... 100'F) if youre bored, the NMC has a temperature sensor you can monitor, its supposed turn the hub warning light red at 50'C (or is that 40?) [snip] |I think, when the room got up to 125 degrees is when they started acting |wierd. Netserver lockups and the like. We did have one catch fire ... P -- O_u \\ U \Beh! \\

> Do any of you apply default filters to all of your users? We apply one to > our "guest" account which disables email, pop, telnet, and ftp. I've often thought about applying filters to *most* customers. If I was applying it to all, in my circumstaces, there would be better places to do it than on the NETServer. What would the implications on the NETServer processor load be if I was to impose, say, a 3 line filter on 40 of the 56 ports? Most traffic would match the 3rd filter line? Regards, Bob Purdon, Technical Manager, Southern Internet Services.

Here is some more info on that multilink freebsd problem.. -- Jan 3 02:17:30 mpd: [nexus] MpWrite: on modem2: seq #1629, 20 bytes Jan 3 02:17:30 mpd: [modem2] sending MP frame: Jan 3 02:17:30 mpd: 3d 40 00 06 5d 0c 79 00 14 05 1b 5d =@..].y....] Jan 3 02:17:30 mpd: fe d0 b6 49 96 50 04 00 00 6c 4a 00 ...I.P...lJ. Jan 3 02:17:30 mpd: 00 c4 82 ... Jan 3 02:17:30 mpd: [modem2] FrameInput: Jan 3 02:17:30 mpd: 3d c0 00 03 4c 2d 7f 00 8a d1 00 0b =...L-...... Jan 3 02:17:30 mpd: be 34 32 36 20 44 61 74 61 20 63 6f .426.Data.co Jan 3 02:17:30 mpd: 6e 6e 65 63 74 69 6f 6e 3a 20 49 6c nnection:.Il Jan 3 02:17:30 mpd: 6c 65 67 61 6c 20 73 65 65 6b 2e 0d legal.seek.. Jan 3 02:17:30 mpd: 0a . I'm not sure exactly what it meant.. or what significance it had. Also, when I tried to receive data, I would get this: Jan 3 02:09:58 graphix mpd: [modem2] ** Bad FCS: 0xac9b != 0xf0b8 This usually ended up with timeouts to LCP keepalives, causing the second (modem2) link to fail, hangup, and re-dial again: Jan 3 02:09:30 graphix mpd: [modem2] carrier detect (CD) signal lost <then hangup> I did not see these things when uploading, however. I experienced upload transfer rates of 4-5k/sec steady which was lower than what I used to get.. but uploads were generally smooth in transfer. Downloads I saw repeated pauses in transfers, continuations, then pauses.. then hangup. -- This is with TCS 3.0 and netserver v3.3.69. - lv

I am trying to get a Cisco 776 to work with our USR TC Hub running TCS3.0. If anyone has had success/failure with this please let me know. I am not even seeing it try to authenticate! If anyone has gotten this to work, and they have the config they used on the client side, that would be most helpfull also. 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 -----------------------------/

I'm trying to find out if the TC supports RFC 2125 BAP/BACP. I know 3Com has BACP on several of the AccessBuilder products, but I don't know about the TC. Is BACP on the TC? If so any restrictions - ie only in the NetServer or only in the HiPer Access? If not, any announced plans for support? -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

Can some one show me how to just grab PPP debug info for one particular port/ip? I am unsure of what port on the hub the call will come in on, but I am sure of the username, ip address etc (its static). I think there is a way to do this using a filter. for now I do set console set debug 0x51 and its way to busy, would like to just look at debug info for a single user/port 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 -----------------------------/

We have a PRI setup and would like to enable DOVBS. How would one go about that with the USR-TC unit? DCP

On Mon, 5 Jan 1998, Jason Kelton wrote: > > Tim, > > Apparently its not fully supported for the TC ENH as its not fully > functional. The only place on TOTALservice its located is in the Netserver > 8/16-I section... NOT the TC area. This is not correct to say. The NETServer manager is funtional - It allows you to download code, setup users etc. The funtions that the NETServer manager includes are those function were supported in 3.2.x NETServer code. The NETServer code has undergone several changes since then and has many new features. Unfortunately the NETServer manager does not include all these features. It is however funtional. krish > > Regards, > > Jason Kelton > 3COM ANZA. > > > > > tdeem2@comsource.net on 01/01/98 04:54:33 > > Please respond to usr-tc@lists.xmission.com > > To: usr-tc@xmission.com > cc: (bcc: Jason Kelton/Support/USR/AU) > Subject: (usr-tc) NetServer Configuration Tool > > > > > > All, > I'm trying to locate the netserver manager software that used to be > available separate form the TCM package. Does anyone happen to know > where it is on the USR site (I've searched totalservice) or have a copy > it they could email me (or have it online for download somewhere). > Any help is appreciated.... > > Thanks, > Timothy > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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. >

I have a Netserver-16/I unit that is rebooting itself every couple of hour. Here is the software information: > Command> version > > U.S. Robotics > Total Control (tm) NETServer 8/16 version 3.2.5.3 > > Build date: Dec 2 1996 > Build time: 12:54:30 To find out what is causing the problem, I connected a pc with a printer to the console port, logged in as !root, and set console. I did not get any indication of why the system is failing. I also have syslogging turned on but no messages relating to the failure are being logged but the syslog is working because I get other messages from the NetServer. Does anyone have any suggestions or if this is not the correct place to ask this question, please point me to the correct list. Thanks. Butch Butch Kemper | Free sound advice available Kemper & Associates Consulting Group | "95% sound and 5% advice" 409-361-2324 | Refunds cheerfully provided

Tim, Apparently its not fully supported for the TC ENH as its not fully functional. The only place on TOTALservice its located is in the Netserver 8/16-I section... NOT the TC area. Regards, Jason Kelton 3COM ANZA. tdeem2@comsource.net on 01/01/98 04:54:33 Please respond to usr-tc@lists.xmission.com cc: (bcc: Jason Kelton/Support/USR/AU) All, I'm trying to locate the netserver manager software that used to be available separate form the TCM package. Does anyone happen to know where it is on the USR site (I've searched totalservice) or have a copy it they could email me (or have it online for download somewhere). Any help is appreciated.... Thanks, Timothy - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.

Once upon a time Lee Kuo shaped the electrons to say... >Are you dialing in or dialing out with the USR-TC unit? If dialing in & >with PRI, then you wouldn't need DOVBS and PRI is already ISDN and you >should just be able to dial using CSD ISDN. I'm not sure how you enable Need technically? No. Need tarriffly? Yes, if you live in the wrong place. The cost for an ISDN DATA call can be several times that of an ISDN DOSBS call in some areas of the country. Bell Atlantics ex-NYNEX turf for example. In some places DATA calls are charged per minute and VOICE calls are NOT. I have many friends back in MA using ISDN who pay less than $200 a month using DOSBS for a 128K nailed up, who would pay closer to $2000 is they used DATA instead. In areas with high ISDN tarriffs DOSBS can be the only economically feasible way to market the service. >DOVS dialing out to CT1. You can't do DOSBS dialing OUT on CT1, but you could support it dialing IN over a CT1. The PM-3 does for example. -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

Four pending PRIs and three overdue evaluation chassis forced me to go online with full 230 channel HiPer HDM/ARC today. Instead of having it at the front of my pool as I once desired, the problems I am experiencing have forced me to put it at the back of my pool. In any case, this is still a third of my total USR dialin in and these problems are affecting me in a big way. I have gone through the proper channels in an attempt to resolve these the "right" way. Apparently USR had a mystery phone number in my records (even though they had called me back at the proper number several times) and thus the case got lost or stalled. I don't know why contact with me was such a problem when I had faxed complete documentation (upon the tech's request) of the bugs to USR. The case number, if anyone cares, is 24271, it mainly dealt with the most serious problems, routing and ISDN-Modem bugs. I was hoping to get those resolved before I started with the others. Here is a complete rundown of problems I am experiencing. I have tried suggested fixes to no avail. I honestly believe that someone needs to dive into the code and get an update out pronto. Bugs in the HiPer HDM Card: 1. Same problems as Quad ISDN-modem. Can't connect to Cisco ISDN adapters. Some other adapters are sketchy at best. 2. V120 enable does not stick in ISDN-modem configuration via TCM. Bugs in the HiPer ARC Card: 1. Routing is not advertised in some cases. This applies to static, framed, and plain "Framed-IP-Address" RADIUS. Eth:1 is 166.70.1.45/26 Cisco is 166.70.1.1/26 Any 166.70.*.* routes do not get advertised, for example: 166.70.32.0/24 166.70.40.0/24 Other routes outside this class B do get advertised, such as: 207.135.128.32/32 198.60.81.0/24 207.135.129.16/28 2. Framed-Routing=None sets routing to RIPv1 3. ARC has mysteriously crashed on me once so far. It has been in full use (230 modems) for less than 12 hours. 4. ARC can not use MPIP with other chassis. 5. RADIUS "Session-Timeout" is either ignored, or "network" activity is defined very loosely. I have not been able to get any sessions to timeout ever. Just about everyone who connects, if they don't manually disconnect will stay connected indefinitely, even though they all have Session-Timeout=600.

Brian Elfert said once upon a time: >On Mon, 5 Jan 1998, Pete Ashdown wrote: > >> 5. RADIUS "Session-Timeout" is either ignored, or "network" activity is >> defined very loosely. I have not been able to get any sessions to >> timeout ever. Just about everyone who connects, if they don't manually >> disconnect will stay connected indefinitely, even though they all have >> Session-Timeout=600. > >Are you sure you're not thinking of Idle-Timeout? > >Session-Timeout is the limit on the total length of a session. My mistake. It is Idle-Timeout that I am talking about. I have been setting this parameter to 600 and not the aforementioned "Session-Timeout". The bug still exists.

On Mon, 5 Jan 1998, Douglas C. Palmer wrote: > We have a PRI setup and would like to enable DOVBS. How would one go about > that with the USR-TC unit? Are you dialing in or dialing out with the USR-TC unit? If dialing in & with PRI, then you wouldn't need DOVBS and PRI is already ISDN and you should just be able to dial using CSD ISDN. I'm not sure how you enable DOVS dialing out to CT1.

On Mon, 5 Jan 1998, Pete Ashdown wrote: > 5. RADIUS "Session-Timeout" is either ignored, or "network" activity is > defined very loosely. I have not been able to get any sessions to > timeout ever. Just about everyone who connects, if they don't manually > disconnect will stay connected indefinitely, even though they all have > Session-Timeout=600. Are you sure you're not thinking of Idle-Timeout? Session-Timeout is the limit on the total length of a session. Brian

Just a minor comment, but I always liked the original representation of the files available for download, all in one box. Now you can no longer see at a glance which versions have been updated. Perhaps that box can be put back and kept up to date, with links to the existing file package breakdown display further down the page? - lv On Mon, 5 Jan 1998, Jason Kelton wrote: > > Tim, > > Apparently its not fully supported for the TC ENH as its not fully > functional. The only place on TOTALservice its located is in the Netserver > 8/16-I section... NOT the TC area. > > Regards, > > Jason Kelton > 3COM ANZA. > > > > > tdeem2@comsource.net on 01/01/98 04:54:33 > > Please respond to usr-tc@lists.xmission.com > > To: usr-tc@xmission.com > cc: (bcc: Jason Kelton/Support/USR/AU) > Subject: (usr-tc) NetServer Configuration Tool > > > > > > All, > I'm trying to locate the netserver manager software that used to be > available separate form the TCM package. Does anyone happen to know > where it is on the USR site (I've searched totalservice) or have a copy > it they could email me (or have it online for download somewhere). > Any help is appreciated.... > > Thanks, > Timothy > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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. >

Yong, Do you have a NETServer in the chassis? If you are trying to use the modems connected via the NETServer - fax is not possible. 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 Tue, 6 Jan 1998 yong_chen@nyed.uscourts.gov wrote: > > I am try to configure the USR-TC analog modems to send FAX. Is that > possible? If that is possible, how can I configure it? Any help would > be greatly appreciated. > Thanks. > > Yong > > Yong Chen > U.S. District Courts, Eastern District of New York > 225 Cadman Plaza E. > Brooklyn, NY 11201 > voice (718) 260-2297 > fax (718) 260-2338 > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Our TC racks have the standard AC->DC power supplies in them, but I noticed some references to the DC power supplies in the documentation on the latest TC documentation cdrom. We're interested in converting our four racks to DC-only, and getting a DC UPS to support them. Does anyone know a manufacturer who makes DC UPSes? All I'd need is a name or a web address. Thanks. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190

I have a PM2 with 30 analog modems...radius running on RedHat Linux. I also have a U.S.Robotics Netserver 8/I-Modem which I use for X2 56K connections. Radius Accounting, authentication, Idle-Timout, etc. work fine for both access servers, with the exception of the Session Timout. In the users file I have:

We use the DC racks from USR with power systems from Argus Technologies. Their website is at www.argus.ca They make nice equipment. At 07:26 AM 1/6/98 -0500, you wrote: >Our TC racks have the standard AC->DC power supplies in them, but I >noticed some references to the DC power supplies in the documentation on >the latest TC documentation cdrom. We're interested in converting our >four racks to DC-only, and getting a DC UPS to support them. > >Does anyone know a manufacturer who makes DC UPSes? All I'd need is a >name or a web address. > >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. > >

How does one go about using the netserver in conjuction with analog lines connected to the nic of DS modems? Is this even poss. If I connect the modems to the nic how do they speak to the bus and therefore the netserver? Do I have to run a seperate com server in order to do this? Terry Kennedy Olypen, Inc.

Krish, My apologies for making my comments unclear. I was insinuating that there are features of the netserver, that are not configurable with the netserver manager, hence your comments qualifying what I was trying to say. Regards, Jason. Please respond to usr-tc@lists.xmission.com cc: usr-tc@lists.xmission.com On Mon, 5 Jan 1998, Jason Kelton wrote: > > Tim, > > Apparently its not fully supported for the TC ENH as its not fully > functional. The only place on TOTALservice its located is in the Netserver > 8/16-I section... NOT the TC area. This is not correct to say. The NETServer manager is funtional - It allows you to download code, setup users etc. The funtions that the NETServer manager includes are those function were supported in 3.2.x NETServer code. The NETServer code has undergone several changes since then and has many new features. Unfortunately the NETServer manager does not include all these features. It is however funtional. krish > > Regards, > > Jason Kelton > 3COM ANZA. > > > > > tdeem2@comsource.net on 01/01/98 04:54:33 > > Please respond to usr-tc@lists.xmission.com > > To: usr-tc@xmission.com > cc: (bcc: Jason Kelton/Support/USR/AU) > Subject: (usr-tc) NetServer Configuration Tool > > > > > > All, > I'm trying to locate the netserver manager software that used to be > available separate form the TCM package. Does anyone happen to know > where it is on the USR site (I've searched totalservice) or have a copy > it they could email me (or have it online for download somewhere). > Any help is appreciated.... > > Thanks, > Timothy > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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.

Thanks on this one! > The modem connection to the NIC for the incoming phone line, and for the > DTE connection are two different connections...they can be connected

I am try to configure the USR-TC analog modems to send FAX. Is that possible? If that is possible, how can I configure it? Any help would be greatly appreciated. Thanks. Yong Yong Chen U.S. District Courts, Eastern District of New York 225 Cadman Plaza E. Brooklyn, NY 11201 voice (718) 260-2297 fax (718) 260-2338

Thus spake Terry Kennedy >How does one go about using the netserver in conjuction >with analog lines connected to the nic of DS modems? >Is this even poss. If I connect the modems to the nic how >do they speak to the bus and therefore the netserver? Do >I have to run a seperate com server in order to do this? The modem connection to the NIC for the incoming phone line, and for the DTE connection are two different connections...they can be connected independantly of each other. Stick the NIC's in there, tell the modem to use the nic for the line source, and do the set modem active thing on the netserver and you'll have what you want. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

I'm experiencing a problem with dual PRIs when they reach capacity (all channels in use). For some reason, calling a number on a pri (or group) when all channels are in use results in a "all circuits busy" message from the telco. Our telco is telling me that this is caused by our equipment not properly rejecting D channel signaling when a B channel is not available. Are they full of it, or is there any truth to this? Thanks ... Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

On a somewhat related note (odd uses of the usr-tc?), is it possible to dial out to a local bank for performing credit card transactions? I know there are some credit card services on the net which allow you to use their secure web server, or PGP c/perl utils but in all cases there is still a 45 second delay or so for the transaction, because they use modems to dial out. If a modem is going to be used at all I would rather have it done locally from my usr-tc, instead of somewhere out on the net. Essentially I plan on writing my own software so what I'm curious about is if this can be done relativly easily, by logging into a system with simple ASCII type commands, or would it require some complex prorietary binary protocol with encryption thats undocumented.. if someone could point me in the right direction I'd appreciate it. :) - lv On Tue, 6 Jan 1998 yong_chen@nyed.uscourts.gov wrote: > > I am try to configure the USR-TC analog modems to send FAX. Is that > possible? If that is possible, how can I configure it? Any help would > be greatly appreciated. > Thanks. > > Yong > > Yong Chen > U.S. District Courts, Eastern District of New York > 225 Cadman Plaza E. > Brooklyn, NY 11201 > voice (718) 260-2297 > fax (718) 260-2338 > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Thus spake System Administrator >I'm experiencing a problem with dual PRIs when they reach capacity (all >channels in use). For some reason, calling a number on a pri (or group) >when all channels are in use results in a "all circuits busy" message from >the telco. Our telco is telling me that this is caused by our equipment >not properly rejecting D channel signaling when a B channel is not >available. Are they full of it, or is there any truth to this? When all B channels are full, it doesn't matter how you have your equipment set, the telco won't even send any signalling on the D channel for your equipment to accept or reject (at least that's our experience) but instead, the telco will generate the "All Circuits Busy" cause code resulting in the fast busy (cause code 58 I believe?). What you can try to test this theory is set the Cause Code settings to "17", which is a "User Busy" and will generate a normal busy signal (I usually set all of them on our equipment to 17's), then if you get a fast busy, you know either something is really broken or the telco is generating it, not you. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

> > Apparently its not fully supported for the TC ENH as its not fully > > functional. > > then and has many new features. Unfortunately the NETServer manager does > not include all these features. It is however funtional. Which makes it 'not *fully* functional' :-) In theory it's fully functional with the features that it supports, but for practical purposes it isn't. Regards, Bob Purdon, Technical Manager, Southern Internet Services.

Our telephone company asks us if we want the "all circuits busy" treatment to be a regular busy signal, a trunk busy (fast busy), or overflow to another group. They set this up when we install new circuits. Our telephone company uses Nortel DMS-100 equipment. Its possible the Lucent switches used in some areas of the U.S. are not as capable, and cannot be configured this way. At 04:05 PM 1/6/98 -0500, you wrote: >I'm experiencing a problem with dual PRIs when they reach capacity (all >channels in use). For some reason, calling a number on a pri (or group) >when all channels are in use results in a "all circuits busy" message from >the telco. Our telco is telling me that this is caused by our equipment >not properly rejecting D channel signaling when a B channel is not >available. Are they full of it, or is there any truth to this? > >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. > >

On Tue, 6 Jan 1998, Jeff Mcadams wrote: > Thus spake System Administrator > >I'm experiencing a problem with dual PRIs when they reach capacity (all > >channels in use). For some reason, calling a number on a pri (or group) > >when all channels are in use results in a "all circuits busy" message from > >the telco. Our telco is telling me that this is caused by our equipment > >not properly rejecting D channel signaling when a B channel is not > >available. Are they full of it, or is there any truth to this? > > When all B channels are full, it doesn't matter how you have your > equipment set, the telco won't even send any signalling on the D channel > for your equipment to accept or reject (at least that's our experience) > but instead, the telco will generate the "All Circuits Busy" cause code > resulting in the fast busy (cause code 58 I believe?). > > What you can try to test this theory is set the Cause Code settings to > "17", which is a "User Busy" and will generate a normal busy signal (I > usually set all of them on our equipment to 17's), then if you get a > fast busy, you know either something is really broken or the telco is > generating it, not you. So does anyone have a complete list of cause codes, or at least partial list? 17 User busy 58 Fast busy If there are enough of them, woudlnt you want to set a different cause code for each condition, to help in determining the problem remotly? 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 | | 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 -----------------------------/

On Tue, 6 Jan 1998, Matthew Opoka wrote: > I just purchased a TC 1706 bundle. > Does 3COM/USR have x86 software for Solaris? No we do not have any Radius software for x86. However you can use Livingston Radius on X86 - no problems > If not are they going to support x86 Solaris? > Can I use my Livingston Radius with the TC and PM2e's? > If so what do I need to do to make it work? > You need to add the NETServer's IP address /name to the client list of the Radius server. krish > Matthew > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Use a 16 meg nonparity, non-edo simm strip. Generic no-name is fine. At 07:58 PM 1/6/98 -0500, you wrote: >Hi, > >Does anyone know the specs on the memory for the NetServer and Network >Management cards in a TCH? Parity or non, EDO, or not, maximum size, etc. > >While I'm asking, anyone have any experience using non-USR memory? > >Thanks, > >Charles > >~~~~~~~~~ ~~~~~~~~~~~ >Charles Sprickman Internet Channel >INCH System Administration Team (212)243-5200 >spork@inch.com access@inch.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. > > 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 ~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~

Hi, Does anyone know the specs on the memory for the NetServer and Network Management cards in a TCH? Parity or non, EDO, or not, maximum size, etc. While I'm asking, anyone have any experience using non-USR memory? Thanks, Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com

Thus spake Brian >So does anyone have a complete list of cause codes, or at least partial >list? I have a pretty comprehensive list, though not in electronic form. :/ I'll see what I can do about getting it typed in...might not be right away though. >17 User busy >58 Fast busy >If there are enough of them, woudlnt you want to set a different cause >code for each condition, to help in determining the problem remotly? Doesn't really help, I've found as basically anything other than a 17 will generate a fast busy signal, making remote problem determination a crap shoot at best. Perhaps if you have a TA that will tell you what cause code is returned, then that would be useful, but dialing with a voice line will all generate the same fast busy. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

I just purchased a TC 1706 bundle. Does 3COM/USR have x86 software for Solaris? If not are they going to support x86 Solaris? Can I use my Livingston Radius with the TC and PM2e's? If so what do I need to do to make it work? Matthew

> Does anyone know the specs on the memory for the NetServer and Network > Management cards in a TCH? Parity or non, EDO, or not, maximum size, etc. > > While I'm asking, anyone have any experience using non-USR memory? Off the shelf, NON-parity, NON-EDO SIMMs. Use them often, and they work just fine. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Exec-PC Vice President, Information Services |

Once upon a time Tatai SV Krishnan shaped the electrons to say... >No we do not have any Radius software for x86. However you can use >Livingston Radius on X86 - no problems >> Can I use my Livingston Radius with the TC and PM2e's? Note only because he also owns PM2s. (You have to own Lucent (Livingston is gone...) HW to legally use our RADIUS.) >You need to add the NETServer's IP address /name to the client list of >the Radius server. He'll probably also want to merge the dictionaries if he wants to use 3Com attribute names. Or just find the equivalent in the default dictionary. Note too that the current 2.0.1 server doesn't support VSAs which 3Com uses for a few things. Our 2.1 server, now in beta, does - along with RADIUS Proxying and a few other things. -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

On Wed, 7 Jan 1998, Eric J. Merkel wrote: > Ok, I have to admit I am new to USR TC hardware, but we just bought > an 1866 bundle. Anyhow, I've got a few questions on what the different > components do. > > First of all, it came with a Netserver card which has an ethernet port > and a couple of WAN connections. The setup of this was relatively simple > and similar to what I have done on PM3's and Netserver 16's. > > Anyhow, my first question is, I am right in assuming that all of the > dialup traffic coming into our 2 channelised T1's goes through the > ethernet port on the netserver card to network? That is correct. > > If so, what is the purpose of the Network Management Card? It also has > an ethernet port, but I am not sure what purpose this card serves? The > documentation is a bit lacking as to the general functionality of the > different devices in the box, hence my dumb questions. The Network Management Card (NMC) is that card that is used to manage the whole chassis, It is a divice that can be accessed via ethernet or via the console port, uses SNMP to communicate, will allow you to download code to all the cards, report problems with cards, send traps and also allows you setup autoresponse etc. > > My second question is, can I telnet to the box and get access to the > Dual T1/PRI card that we have in the box? I am accustomed to doing > a "sh line0" on a PM3 to see if a span is up and working. So far all > I've been able to do is telnet to the terminal server card and the T1 > line status is nowhere to be found. This is really handy for looking at > our remote POP's and would like to be able to figure out how to do it > on this box as well...I really don't want to have to drive out to the > location to look at the status lights or hook up a serial connection. No - not directly from the NETServer. 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 > > Thank for responding to my questions!!! > 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. > ============================================================================= > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

72 pin, 60-70ns, fast page, no parity, gold connector, single sided The last one got me. If you use double sided SIMMs, they may not fit on the Netserver card. At 10:26 PM 1/6/98 -0600, you wrote: >> Does anyone know the specs on the memory for the NetServer and Network >> Management cards in a TCH? Parity or non, EDO, or not, maximum size, etc. >> >> While I'm asking, anyone have any experience using non-USR memory? > >Off the shelf, NON-parity, NON-EDO SIMMs. Use them often, and they work >just fine. > >-------------------------------------------------------------------------- >| Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | >| Exec-PC Vice President, Information 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. > >

We are running a USR TC Enterprise hub with a high speed WAN or Ethernet NIC controller card which gives the following description with the version command: U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.5.34 Build date: Jun 19 1997 Build time: 14:02:28 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Enhanced Licensed for 60 ports. We run it on an ehternet segment with 2 Livingston portmasters doing OSPF and a UNIX router running gated 3.5.8 (which is the exit point). The rest of our routing is done using OSPF, and we want to have a routing protocol on the TC rack that supports full variable subnet masking and nexthop information as we have all sorts of subnets everywhere (with differing sizes on the dial-up routes) and to save packets from having to traverse the ethernet segment twice. There are 3 things I have discovered with the RIP V2 code on the above when trying to get it to function with gated: 1) It does NOT do RIPv2 packets via multicast as per the RFC. This prevents me form exporting various routes from OSPF into RIPv2 that would be beneficial to the operation of the TC rack 2) It does full RIPv2 packets on the ethernet IP broadcast address, rather than RIPv1 compatible RIPv2 packets. 3) It takes no notice of the nexthop information in the incoming RIPv2 packets. We did manage to get it funcitoning after a fashion with only a default route pointing at the gated box, and gated listening to the RIPv2 from the TC rack. The time taken for routes to be deleted from the TC rack when a customer hangs up due to RIPv2 is also a minor problem. Can anyone shed some light on what USR are doing with their RIPv2 implementation? Thanks, Chris.

On Tue, 6 Jan 1998, Brian wrote: > On Tue, 6 Jan 1998, Jeff Mcadams wrote: > > > Thus spake System Administrator > > >I'm experiencing a problem with dual PRIs when they reach capacity (all > > >channels in use). For some reason, calling a number on a pri (or group) > > >when all channels are in use results in a "all circuits busy" message from > > >the telco. Our telco is telling me that this is caused by our equipment > > >not properly rejecting D channel signaling when a B channel is not > > >available. Are they full of it, or is there any truth to this? > > > > When all B channels are full, it doesn't matter how you have your > > equipment set, the telco won't even send any signalling on the D channel > > for your equipment to accept or reject (at least that's our experience) > > but instead, the telco will generate the "All Circuits Busy" cause code > > resulting in the fast busy (cause code 58 I believe?). > > > > What you can try to test this theory is set the Cause Code settings to > > "17", which is a "User Busy" and will generate a normal busy signal (I > > usually set all of them on our equipment to 17's), then if you get a > > fast busy, you know either something is really broken or the telco is > > generating it, not you. > > So does anyone have a complete list of cause codes, or at least partial > list? I often use the list on ascend's web site, but it's not too helpful in this situation. I *did* discover that changing "Analog Connection Blocked" and "Digital Connection Blocked" to cause code 17 instead of 58 resulted in a proper busy tone rather than re-order (fast busy). This means that my telco *is* attempting to signal new calls even though all B channels are in use. On another interesting note, if I leave the cause code for "Specific B channel blocked" set to 58 _AND_ I take one of the PRI timeslots out of service, the telco doesn't seem to be able to hunt inbound calls to any other PRIs in the hunt-group (re-order tone is signaled); however if I set this to cause code 17, hunting works properly even with a timeslot out-of-service. Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

On Wed, 7 Jan 1998, Jeff Mcadams wrote: > Heh...yup...same copy I have...printed out. I pulled it off dejanews > when I got it. Its still accessible, do a search for: > ISDN cause codes comp.dcom.isdn document list > on the old database and you'll find it. I found it on Cisco's web site: http://www-europe.cisco.com/univercd/data/doc/software/11_1/dcr/disdn.htm#REF27122 -CH

Thus spake Phil Dye >If you can find it, then <DG0BuK.FKG@world.std.com> from comp.dcom.isdn >has a comprehensive list. However, the article dates from October 95, >so I'm having trouble getting hold of it (other than the printed out >copy I have ;) Heh...yup...same copy I have...printed out. I pulled it off dejanews when I got it. Its still accessible, do a search for: ISDN cause codes comp.dcom.isdn document list on the old database and you'll find it. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

At 5:21 AM -0800 1/7/98, MegaZone wrote: >Once upon a time Tatai SV Krishnan shaped the electrons to say... >>No we do not have any Radius software for x86. However you can use >>Livingston Radius on X86 - no problems >>> Can I use my Livingston Radius with the TC and PM2e's? > >Note only because he also owns PM2s. (You have to own Lucent (Livingston >is gone...) HW to legally use our RADIUS.) This is only true for 2.X versions of RADIUS, no? I was under the impression that it was still okay to use something like 1.16 without owning any Livingston products. Other versions of RADIUS (Merit, Cistron...) should also work okay if you get them to compile for you. Jordyn |----------------------------------------------------------------| |Jordyn A. Buchanan mailto:jordyn@bestweb.net | |Bestweb Corporation http://www.bestweb.net | |Senior System Administrator +1.914.271.4500 | |----------------------------------------------------------------|

We have USR-TC with a Dual Pri card and 6 quad modem cards via the Netserver. I downloaded "NCSI Client 95" software and installed it in my desktop which to add a NCSI Shared Port in Com 4. Now, I try to use Hyperterminal to dialout(use u.s. robotic "Courier V Everthing" modem via NCSI Shared Port) in Windows 95 enviroment. The following is the message I got: "Another program is using the selected Telephony device. Try again after the program completes." What I did wrong and how to fix this problem? By the way, our USR-TC is a new installation unit. It works fine on dialin. Any help would be greatly appreciated. Yong. Yong Chen U.S. District Courts EDNY Voice (718) 260-2297

> > > > So does anyone have a complete list of cause codes, or at least partial > > list? > > I often use the list on ascend's web site, but it's not too helpful in > this situation. > > I *did* discover that changing "Analog Connection Blocked" and "Digital > Connection Blocked" to cause code 17 instead of 58 resulted in a proper > busy tone rather than re-order (fast busy). This means that my telco *is* > attempting to signal new calls even though all B channels are in use. > > On another interesting note, if I leave the cause code for "Specific B > channel blocked" set to 58 _AND_ I take one of the PRI timeslots out of > service, the telco doesn't seem to be able to hunt inbound calls to any > other PRIs in the hunt-group (re-order tone is signaled); however if I set > this to cause code 17, hunting works properly even with a timeslot > out-of-service. Everything here is set for a cause code of 58, I am talking about the span level cause codes now. This is how they come default. I take span's of service usually by just unplugging them, and this allows hunting to continue as it should. Its easier for me to do that, with the equiptment that is local. I would like to, at a minimum, set traps for ANY blocking that occurs (analog, digital, no modems available, no isdn available, specific b channel blocking). This is usually are biggest problem. If its a busy, and its generated local by our equiptment, we want a log showing something is amiss. Anyone doing this? Can you tell me what to set the traps on, and possible recommend a good system on Unix to capture the traps? Thanks. Brian > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > /-------------------------- 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 -----------------------------/

Brian said; >On Tue, 6 Jan 1998, Jeff Mcadams wrote: > >So does anyone have a complete list of cause codes, or at least partial >list? > >17 User busy >58 Fast busy If you can find it, then <DG0BuK.FKG@world.std.com> from comp.dcom.isdn has a comprehensive list. However, the article dates from October 95, so I'm having trouble getting hold of it (other than the printed out copy I have ;) -- Phil Dye | Work: pmd@tcp.net.uk Network Manager | Play: phil@lart.ing.co.uk Total Connectivity Providers | Consider myself properly disclaimed "The nice thing about standards is that there are so many to choose from" -anon

Ok, I have to admit I am new to USR TC hardware, but we just bought an 1866 bundle. Anyhow, I've got a few questions on what the different components do. First of all, it came with a Netserver card which has an ethernet port and a couple of WAN connections. The setup of this was relatively simple and similar to what I have done on PM3's and Netserver 16's. Anyhow, my first question is, I am right in assuming that all of the dialup traffic coming into our 2 channelised T1's goes through the ethernet port on the netserver card to network? If so, what is the purpose of the Network Management Card? It also has an ethernet port, but I am not sure what purpose this card serves? The documentation is a bit lacking as to the general functionality of the different devices in the box, hence my dumb questions. My second question is, can I telnet to the box and get access to the Dual T1/PRI card that we have in the box? I am accustomed to doing a "sh line0" on a PM3 to see if a span is up and working. So far all I've been able to do is telnet to the terminal server card and the T1 line status is nowhere to be found. This is really handy for looking at our remote POP's and would like to be able to figure out how to do it on this box as well...I really don't want to have to drive out to the location to look at the status lights or hook up a serial connection. Thank for responding to my questions!!! 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. =============================================================================

> > My second question is, can I telnet to the box and get access to the > Dual T1/PRI card that we have in the box? I am accustomed to doing > a "sh line0" on a PM3 to see if a span is up and working. So far all > I've been able to do is telnet to the terminal server card and the T1 > line status is nowhere to be found. This is really handy for looking at > our remote POP's and would like to be able to figure out how to do it > on this box as well...I really don't want to have to drive out to the > location to look at the status lights or hook up a serial connection. > > Thank for responding to my questions!!! > Eric > What we do here, is we have a multiport serial card, such as a Cyclades, Digiboard, or Rocketport. This is hooked into a Linux box, with serial lines running to each device (netserver, pri, hdm's, etc). We also hang a modem off the serial board. For administration, you can simply write a minirc.xxx file, so when you type "pri1" or "netserver2" etc, it takes you into the appropriate card. This is especially good for when all hell breaks loose, you cant tcp/ip into the cards, you cant dial into the tc hub, etc. You can always fire up a term prog, dial into the linux box, get to a shell, and serial over to your hardware. With a remote pop this is a good idea also. You can buy cheap Boca 2016 boards and they will cost you next to nothing, and work fine for simple admin stuff. > > > ============================================================================= > 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 | | 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 -----------------------------/

On Wed, 7 Jan 1998 15:23:47 -0600 (CST), Brian <signal@shreve.net> wrote: >Anyone doing this? Can you tell me what to set the traps on, and possible >recommend a good system on Unix to capture the traps? I'm using ucd-snmp3.3.pre5 to capture traps from a number of devices. It's at ftp://ftp.ece.ucdavis.edu/pub/snmp/ There was something I had changed in the code to get it to recognize USR's traps, but I think that got fixed in ucd-snmp3.3.1. I've got a perl script that parses the output of snmptrapd and adds the traps into a mSQL database. Is there a file archive I can send the script to? Russ

What I did was setup a port as a "login" port, setup a pool to connect to that port, and added a temporary user to the local NETServer table that would telnet to a Unix box. [ren-usr> set s52 login [ren-usr> add user testuser [ren-usr> set user testuser host myhost [ren-usr> set user testuser password mypassword [ren-usr> set user service telnet So, I dial in to that port, manually type in the username/password, get connected to the Unix box, login in to that box, and finally I run "ppp -direct" which creates a PPP session. Essentially, I moved the PPP processing to another box. This brought Quake pings back down to a normal 200-300 range. Using the PPP in the NETServer (PPP in modem) gives me 800-1200 Quake pings. Question: Is there a way to force the NETServer to telnet directly to a host without requiring the user to authenticate with the NETServer? i.e. User connects, User gets telnetted to a host, and the host authenticates via PAP/CHAP/etc Russ

On Tue, 6 Jan 1998 16:06:28 -0500 (EST), Laszlo Vecsey <master@internexus.net> wrote: >On a somewhat related note (odd uses of the usr-tc?), is it possible to >dial out to a local bank for performing credit card transactions? > >I know there are some credit card services on the net which allow you to >use their secure web server, or PGP c/perl utils but in all cases there is >still a 45 second delay or so for the transaction, because they use modems >to dial out. If a modem is going to be used at all I would rather have it >done locally from my usr-tc, instead of somewhere out on the net. > >Essentially I plan on writing my own software so what I'm curious about is >if this can be done relativly easily, by logging into a system with simple >ASCII type commands, or would it require some complex prorietary binary >protocol with encryption thats undocumented.. if someone could point >me in the right direction I'd appreciate it. :) It should be possible as long as you have software that can interface with the credit bureau you are dealing with. Setup a port as a dialout device that you can connect to via telnet. Point your software at that port, have it dial out using normal AT commands, and let the credit verification software do it's thing. You'll want to restrict who can connect to that dialout port, of course. Russ

Doesn't the 1866 bundle or any TC chassis come with the TCM sotfware? It didn't with mine and I don't have access to download it off of totalservice.usr.com. How can I get ahold of a copy of it? I've got a valid ID from our Netserver support agreements as well as a support agreement we bought with this bundle. Shouldn't I be able to access this software from your website? Eric At 06:35 AM 1/7/98 -0600, you wrote: >On Wed, 7 Jan 1998, Eric J. Merkel wrote: > >> Ok, I have to admit I am new to USR TC hardware, but we just bought >> an 1866 bundle. Anyhow, I've got a few questions on what the different >> components do. >> >> First of all, it came with a Netserver card which has an ethernet port >> and a couple of WAN connections. The setup of this was relatively simple >> and similar to what I have done on PM3's and Netserver 16's. >> >> Anyhow, my first question is, I am right in assuming that all of the >> dialup traffic coming into our 2 channelised T1's goes through the >> ethernet port on the netserver card to network? >That is correct. > >> >> If so, what is the purpose of the Network Management Card? It also has >> an ethernet port, but I am not sure what purpose this card serves? The >> documentation is a bit lacking as to the general functionality of the >> different devices in the box, hence my dumb questions. > >The Network Management Card (NMC) is that card that is used to manage the >whole chassis, It is a divice that can be accessed via ethernet or via >the console port, uses SNMP to communicate, will allow you to download >code to all the cards, report problems with cards, send traps and also >allows you setup autoresponse etc. > > >> >> My second question is, can I telnet to the box and get access to the >> Dual T1/PRI card that we have in the box? I am accustomed to doing >> a "sh line0" on a PM3 to see if a span is up and working. So far all >> I've been able to do is telnet to the terminal server card and the T1 >> line status is nowhere to be found. This is really handy for looking at >> our remote POP's and would like to be able to figure out how to do it >> on this box as well...I really don't want to have to drive out to the >> location to look at the status lights or hook up a serial connection. > >No - not directly from the NETServer. > >krish ============================================================================= 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. =============================================================================

At 10:57 PM 1/7/98 -0600, Russ Panula wrote: >What I did was setup a port as a "login" port, setup a pool to connect >to that port, and added a temporary user to the local NETServer table >that would telnet to a Unix box. > >[ren-usr> set s52 login >[ren-usr> add user testuser >[ren-usr> set user testuser host myhost >[ren-usr> set user testuser password mypassword >[ren-usr> set user service telnet > >So, I dial in to that port, manually type in the username/password, >get connected to the Unix box, login in to that box, and finally I run >"ppp -direct" which creates a PPP session. > >Essentially, I moved the PPP processing to another box. Very clever of you... What made you think of trying this?.. >This brought Quake pings back down to a normal 200-300 range. Using >the PPP in the NETServer (PPP in modem) gives me 800-1200 Quake pings. Really.. I really find this interesting. Since the same amount of aggregate bandwidth is pushed through the netserver either way, maybe there is enough "horsepower" in the netserver afterall. That is unless PPP is a "thick" layer that requires alot of processing.. I wouldn't think so.. I was beginning to wonder if a code fix was even possible. >Question: Is there a way to force the NETServer to telnet directly to >a host without requiring the user to authenticate with the NETServer? >i.e. User connects, User gets telnetted to a host, and the host >authenticates via PAP/CHAP/etc > >Russ > Sure would be cool.. If you get this working in a pretty much transparent fashion, please let us know. I'd almost give up PAP/CHAP on one box to get rid of Quake lag anyway.. Almost.... 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

At 10:57 PM 1/7/98 -0600, Russ Panula wrote: >Question: Is there a way to force the NETServer to telnet directly to >a host without requiring the user to authenticate with the NETServer? >i.e. User connects, User gets telnetted to a host, and the host >authenticates via PAP/CHAP/etc Would it be possible for most users to continue using PAP and authenticate the "normal" way through the netserver. Then have quake users setup a special connect script that authenticates through NetServer and initiates PAP on the unix box? Even with two passwords, it should work?.. And also, could this sort of thing lead to any problem situations with IP routing?.. 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

On Thu, 08 Jan 1998 00:11:39 -0600, Allen Marsalis <am@shreve.net> wrote: >>Essentially, I moved the PPP processing to another box. > >Very clever of you... What made you think of trying this?.. Vacation. My mind thinks clearer while riding up a chairlift. >>This brought Quake pings back down to a normal 200-300 range. Using >>the PPP in the NETServer (PPP in modem) gives me 800-1200 Quake pings. > >Really.. I really find this interesting. Since the same amount of >aggregate bandwidth is pushed through the netserver either way, maybe >there is enough "horsepower" in the netserver afterall. That is unless >PPP is a "thick" layer that requires alot of processing.. I wouldn't >think so.. I was beginning to wonder if a code fix was even possible. Well, horsepower or not, something is amiss with the way the NETServer is handling UDP packets. I guess I tend towards the NETServer not having enough CPU to process all those PPP sessions, but that's the "A 486 is junk" in me talking. Russ

On Thu, 08 Jan 1998 00:40:31 -0600, Allen Marsalis <am@shreve.net> wrote: >At 10:57 PM 1/7/98 -0600, Russ Panula wrote: >>Question: Is there a way to force the NETServer to telnet directly to >>a host without requiring the user to authenticate with the NETServer? >>i.e. User connects, User gets telnetted to a host, and the host >>authenticates via PAP/CHAP/etc > >Would it be possible for most users to continue using PAP and authenticate >the "normal" way through the netserver. Then have quake users setup a special >connect script that authenticates through NetServer and initiates PAP on >the unix box? Even with two passwords, it should work?.. And also, could >this sort of thing lead to any problem situations with IP routing?.. Well, I'm hoping there is some way I can force the NETServer to skip authentication and just telnet each incoming connection to a host I specify. Then from that host I can use mgetty to handle PAP/CHAP authentication and setup the PPP session. Essentially making it so the end user has to make zero changes on their end. IP assignment and routing would be handled by the host providing the PPP sessions, so you'd have to implement whatever routing scheme you're using on that box. Heck, one major side benefit is you can use gated and get OSPF support on your dialups. My mouth is watering already. Russ

Once upon a time Jordyn A. Buchanan shaped the electrons to say... >This is only true for 2.X versions of RADIUS, no? I was under the >impression that it was still okay to use something like 1.16 without owning Yes, 2.0 and up. Note that a major security bulletin went out on 1.16 - it has a known security hole. There is a patched version out on our site now. But this also effects any servers based off of 1.16 - ESVA for example. >any Livingston products. Other versions of RADIUS (Merit, Cistron...) >should also work okay if you get them to compile for you. Personally I recommend Cistron to folks. -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

At 01:32 AM 1/8/98 -0600, Russ Panula wrote: >On Thu, 08 Jan 1998 00:40:31 -0600, Allen Marsalis <am@shreve.net> >wrote: > >>At 10:57 PM 1/7/98 -0600, Russ Panula wrote: >>>Question: Is there a way to force the NETServer to telnet directly to >>>a host without requiring the user to authenticate with the NETServer? >>>i.e. User connects, User gets telnetted to a host, and the host >>>authenticates via PAP/CHAP/etc >> >>Would it be possible for most users to continue using PAP and authenticate >>the "normal" way through the netserver. Then have quake users setup a special >>connect script that authenticates through NetServer and initiates PAP on >>the unix box? Even with two passwords, it should work?.. And also, could >>this sort of thing lead to any problem situations with IP routing?.. > >Well, I'm hoping there is some way I can force the NETServer to skip >authentication and just telnet each incoming connection to a host I >specify. Yeah that would be better to make it transparent to current users configs but _if_ there is no way to do it, then this idea is still a good one. Only the quake users would have to make changes which would be perfectly fine for those wishing to eliminate lag.. And I was thinking it might be better to split things up but maybe not. If a single unix box handled all PPP for all hubs, it could get a little busy on 10T.. >Then from that host I can use mgetty to handle PAP/CHAP authentication >and setup the PPP session. Essentially making it so the end user has >to make zero changes on their end. Understood. I think that's really smart. Simple and resourceful. But I'm no veteran at TC and maybe others have done this to go around bugs n stuff in the early netserver code. (there were bugs back then huh?) >IP assignment and routing would be handled by the host providing the >PPP sessions, so you'd have to implement whatever routing scheme >you're using on that box. Heck, one major side benefit is you can use >gated and get OSPF support on your dialups. My mouth is watering >already. > >Russ I'm with ya.. OSPF huh.. Good reason to route *all* PPP sessons to linux instead of spitting them up. This solution seems to have alot of promise and takes some of the "proprietary wait" out of the TC picture.. Not that I mind being at someones total mercy mind you.... Allen

On Thu, 8 Jan 1998, Russ Panula wrote: > Well, horsepower or not, something is amiss with the way the NETServer > is handling UDP packets. I guess I tend towards the NETServer not > having enough CPU to process all those PPP sessions, but that's the "A > 486 is junk" in me talking. If the 486 isn't fast enough, how is it that Livingston can do thirty modems on a 386 just fine, and 60 modems on a 486? Remember, the Netserver is basically a faster version of a Livingston PM2. The code in the Netserver is modified Livingston code, of course, now so bloated by USR that the code requires 16 megs of RAM. Brian

Why would all of my modems suddenly go AIP in the Netserver? I'm running 3.5.34. Having the modems go AIP wouldn't have been so bad, but the T1 NAC didn't autobusy the two spans as it should have. I could have lived being down 48 modems in the middle of the night. Modems were answering with dead air. Reactivating the modems and rebooting the Netserver seems to have fixed things. Brian

Whenever I run Zmodem through a terminal session on the Netserver, my connection is lost as soon as sz is done. Any ideas to fix this? Sz works fine through a PM2. Brian

At 11:04 PM 1/7/98 -0600, you wrote: >>On a somewhat related note (odd uses of the usr-tc?), is it possible to >>dial out to a local bank for performing credit card transactions? >> >>I know there are some credit card services on the net which allow you to >>use their secure web server, or PGP c/perl utils but in all cases there is >>still a 45 second delay or so for the transaction, because they use modems >>to dial out. If a modem is going to be used at all I would rather have it >>done locally from my usr-tc, instead of somewhere out on the net. >> >>Essentially I plan on writing my own software so what I'm curious about is >>if this can be done relativly easily, by logging into a system with simple >>ASCII type commands, or would it require some complex prorietary binary >>protocol with encryption thats undocumented.. if someone could point >>me in the right direction I'd appreciate it. :) We use a program called ICVERIFY to handle credit card transactions. I'm not sure it makes much sense to reinvent the wheel. You might want to contact them to see if you can fit it in to your pricing model. It handles things very well with very little maintenance and has very good reporting options. The thing I really like about the software is that if things get really busy it will actually hold the line and process transactions one after the other without redialing. If you have enough business to keep the thing full you can get blazingly fast authorizations.

On Wed, 7 Jan 1998, Russ Panula wrote: > On Wed, 7 Jan 1998 15:23:47 -0600 (CST), Brian <signal@shreve.net> > wrote: > > >Anyone doing this? Can you tell me what to set the traps on, and possible > >recommend a good system on Unix to capture the traps? > > I'm using ucd-snmp3.3.pre5 to capture traps from a number of devices. > > It's at ftp://ftp.ece.ucdavis.edu/pub/snmp/ > > There was something I had changed in the code to get it to recognize > USR's traps, but I think that got fixed in ucd-snmp3.3.1. > > I've got a perl script that parses the output of snmptrapd and adds > the traps into a mSQL database. Is there a file archive I can send > the script to? > > Russ Russ, No archive as of yet, but please send me the script, and I will go get ucd, I am very interested. Thanks. Pete, do you have a TC ftp site over at xmission? Since the list/list archive is there, it would save people from having to remember more than one place. I have alot of stuff I could upload too. 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 -----------------------------/

On Thu, 8 Jan 1998, Brian Elfert wrote: > Whenever I run Zmodem through a terminal session on the Netserver, my > connection is lost as soon as sz is done. > > Any ideas to fix this? Sz works fine through a PM2. > I believe this is a known problem. What you need to do is try to establish a Rlogin connection instead of a telnet session, this helps. 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 | | 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 -----------------------------/

On Wed, 7 Jan 1998, Russ Panula wrote: > What I did was setup a port as a "login" port, setup a pool to connect > to that port, and added a temporary user to the local NETServer table > that would telnet to a Unix box. > > [ren-usr> set s52 login > [ren-usr> add user testuser > [ren-usr> set user testuser host myhost > [ren-usr> set user testuser password mypassword > [ren-usr> set user service telnet > > So, I dial in to that port, manually type in the username/password, > get connected to the Unix box, login in to that box, and finally I run > "ppp -direct" which creates a PPP session. > > Essentially, I moved the PPP processing to another box. > > This brought Quake pings back down to a normal 200-300 range. Using > the PPP in the NETServer (PPP in modem) gives me 800-1200 Quake pings. > > Question: Is there a way to force the NETServer to telnet directly to > a host without requiring the user to authenticate with the NETServer? > i.e. User connects, User gets telnetted to a host, and the host > authenticates via PAP/CHAP/etc I am sure there is. The problem lies in how many users are you going to have the linux box handle, how many hubs of PPP sessions will the linux box do? Has RADIUS support been hacked into mgetty? This would be important. Also, I am not sure how robust "Telnet" of the netserver is, since many times you here of zmodem xfers being killed. Perhaps rlogin would be better? If nothing else, it would be good for USR to use to use this to "test" how robust there PPP is. They should get it so that sessions thru pppd and sessions thru the netserver are the same latency. > > Russ > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Thu, 8 Jan 1998, Brian wrote: > On Thu, 8 Jan 1998, Brian Elfert wrote: > > > Whenever I run Zmodem through a terminal session on the Netserver, my > > connection is lost as soon as sz is done. > > > > Any ideas to fix this? Sz works fine through a PM2. > > > > I believe this is a known problem. What you need to do is try to > establish a Rlogin connection instead of a telnet session, this helps. We're already using rlogin, so that won't help unfortunately. Brian

Brian Elfert said once upon a time: >On Thu, 8 Jan 1998, Brian wrote: > >> On Thu, 8 Jan 1998, Brian Elfert wrote: >> >> > Whenever I run Zmodem through a terminal session on the Netserver, my >> > connection is lost as soon as sz is done. >> > >> > Any ideas to fix this? Sz works fine through a PM2. >> > >> >> I believe this is a known problem. What you need to do is try to >> establish a Rlogin connection instead of a telnet session, this helps. > >We're already using rlogin, so that won't help unfortunately. This is an old problem. I ran into it last Spring, and I believe David Bolen was able to replicate it and describe the bug to USR's engineers. I don't know if recent code fixes it or not. What I do know is that you can fix the problem with no side-effects by commenting out a line of code in rbsb.c: (void) ioctl(Tty, TCFLSH, 1); /* Flush input queue */ Hope that helps. You can test this easily with the "sz -T" command, which would hang up after the test as well.

Brian said once upon a time: >Also, I am not sure how robust "Telnet" of the netserver is, since many >times you here of zmodem xfers being killed. Perhaps rlogin would be >better? I seem to recall that the problem only applies to rlogin, not telnet.

What method do those of you using Netservers use to disconnect a user after a certain length of time (regardless of idle state). I am using Radius on RedHat Linux 4.1. I use Radius's Session-Timout on my PM2. Curt

Russ Panula said once upon a time: >So, I dial in to that port, manually type in the username/password, >get connected to the Unix box, login in to that box, and finally I run >"ppp -direct" which creates a PPP session. > >Essentially, I moved the PPP processing to another box. > >This brought Quake pings back down to a normal 200-300 range. Using >the PPP in the NETServer (PPP in modem) gives me 800-1200 Quake pings. This is an interesting idea, but I wonder how many PPP sessions you could run on a single Linux box. Note that if you're the only one connected with PPP on a Netserver, your Quake pings are fairly peppy.

On Thu, 8 Jan 1998, Pete Ashdown wrote: > This is an old problem. I ran into it last Spring, and I believe David > Bolen was able to replicate it and describe the bug to USR's engineers. I > don't know if recent code fixes it or not. Actually, I just dialed into a unit running TCS 3.0, and it seems to work just fine. No hangup after an sz session. With TCS 2.5.1, it consistently hangs up after a zmodem session. I'll have to busy out my chassis and upgrade the Netserver code. Brian

Can you write the netservers settings out to a text file? If so how is that done? I posted this question a few days ago and don't know if it got lost or what, so please excuss me if someone has already addressed this Terry Kennedy .

At 07:49 AM 1/8/98 -0600, Brian wrote: >On Wed, 7 Jan 1998, Russ Panula wrote: >> Question: Is there a way to force the NETServer to telnet directly to >> a host without requiring the user to authenticate with the NETServer? >> i.e. User connects, User gets telnetted to a host, and the host >> authenticates via PAP/CHAP/etc > >I am sure there is. > >The problem lies in how many users are you going to have the linux box >handle, how many hubs of PPP sessions will the linux box do? With modems hanging off 10T, I would think that 3 mbit is all that's going to happen. That's still alot of modem sessions. But 100T would be the way to go with this. The bottleneck would be in the linux box and I'm sure it can handle many hundreds anyway.. Good enough for us. >Has RADIUS support been hacked into mgetty? This would be important. > >Also, I am not sure how robust "Telnet" of the netserver is, since many >times you here of zmodem xfers being killed. Perhaps rlogin would be >better? These are good points worth investigating. >If nothing else, it would be good for USR to use to use this to "test" how >robust there PPP is. They should get it so that sessions thru pppd and >sessions thru the netserver are the same latency. > I'm looking for a solution to "proprietary lag". It's kinda like quake lag only ping times are 6 to 12 months ;) It would be nice to have OSPF and no quake lag which is my reason for wanting to try this. What's good for USR and their testing is another matter entirely that I just as soon not comment on.. Allen

At 12:59 PM 1/8/98 -0800, you wrote: > >Can you write the netservers settings out to a text file? >If so how is that done? >I posted this question a few days ago and don't know >if it got lost or what, so please excuss me if someone has >already addressed this > At this time there is no way to write the config out into a text file. -M `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)

On Thu, 8 Jan 1998, Terry Kennedy wrote: > > Can you write the netservers settings out to a text file? no > If so how is that done? you have to manually type them into a text file the first time you configure. Then you can use that file from then on out and paste it into the netserver to ever reconfigure. > I posted this question a few days ago and don't know > if it got lost or what, so please excuss me if someone has > already addressed this > > Terry Kennedy . > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

rpanula@dacmail.net (Russ Panula) writes: > Question: Is there a way to force the NETServer to telnet directly to > a host without requiring the user to authenticate with the NETServer? > i.e. User connects, User gets telnetted to a host, and the host > authenticates via PAP/CHAP/etc I believe that if you turn off the security option on your ports, and set up a global default host list (or a per-port default host list) that immediately upon receiving a connection the user will be connected (using the default protocol defined for the port) to the global or port host definition. If that doesn't work, you could always enable RADIUS authentication prior to the login prompt: > set radius prompt none username XXXX password YYYY what this does is that upon receiving a new connection, after generating the banner, but before displaying "login:", the NETServer will generate a RADIUS request. Because RADIUS requires username and password attributes, the NETServer will fill in the request using the XXXX and YYYY in the above command. You could then define that user to do whatever you wanted in your authentication system. Either of these options will disable the ability to use that NETServer for other authentications or services. I would also suggest using Netdata (raw TCP) rather than telnet to ensure that you don't have to worry about being 8-bit clean and/or quoting stuff. Of course, if the application listening on the other end is expecting the telnet protocol you may get a few characters of noise due to the options it may output but that shouldn't stop it from working. If you're using at least 3.5.34 (the first release I noticed it in), there is a new delay setting for controlling prompting. If in addition to the "set radius" command above you also do: > set radius prompt delay ## (## = seconds) then the NETServer will wait ## seconds before auto-generating the "fake" login RADIUS request. During that time it will detect PPP logins. Depending on your setup, this may allow you to service other customers, while having Quake customers just delay a few seconds before hitting the target host. Of course, since the Quake customers are probably doing PAP/CHAP logins too, it might not be quite right for this case (unless you instruct the Quake users to wait for something that your separate host outputs before starting their PPP negotation). -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/

On Fri, 9 Jan 1998, Douglas O'Flaherty wrote: > > I have been asked to "help" another company get their netserver to > authenticate using the USR/3com RADIUS server against the local password > file on a Unix system. There is 1 big catch. I'm 1000 miles from the > console and can't get x to display back at this moment. > > I got them to tak to each other. I can have a user appear to authenticate, > and can use an interactive terminal window to log into the netserver > using the password from the unix server. > This was accomplished by moving my db to the postgres db and using the > postgres sql calls (and sautil_pg -v -d) to really look at the database > and determine what is munged. > > BUT.... > > I can't get windows 95 users to login. It will work if we setup the user > on the netserver card. > > I've been through 3 days of 3com support (different very long rant) and > gotten nothing! > > I think I am looking for either > 1) the proper "set" command on the Netserver which I don't know > and/or > 2) a set of sample user and system config files which are functional. > > This has never been a problem on other tcm systems with which I have worked. > And I am stumped. Tech support has been useless. > First make sure the NETServer has the following settings set pap on set chapfst off This will tell the NETServer to do PAP and first start doing PAP instead of CHAP. You are using system password and if you use system password you will not be able to do chap. Make sure that the win95 machine that you are using in dialup networking is not using encrypted password, and if you have software compression on turn it off and try. As far as the Radius is concerened the user should be a general netserver user and should be using system password. Also if you are not setting an IP address you should set the assigned ip address pool on the NETServer set assigned < ip address pool > This should work krish > doug > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

I have a new chassis with tcs2.5.1 on it. The card for the span lines is flashed with the pri code right now. My question is can I flash this card with the t1 code and have it work? Thanks, Pat

On Thu, 8 Jan 1998 pat@coffey.com wrote: > I have a new chassis with tcs2.5.1 on it. The card for the span lines > is flashed with the pri code right now. My question is can I flash > this card with the t1 code and have it work? > > Thanks, > Pat > > yep, just flash it T1. Ours came T1, and we flashed PRI. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

Thanks, for the info Brian. I'll br flashing it in the morning. Pat On Thu, 8 Jan 1998, Brian wrote: > On Thu, 8 Jan 1998 pat@coffey.com wrote: > > > I have a new chassis with tcs2.5.1 on it. The card for the span lines > > is flashed with the pri code right now. My question is can I flash > > this card with the t1 code and have it work? > > > > Thanks, > > Pat > > > > > > yep, just flash it T1. Ours came T1, and we flashed PRI. > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. >

hmm, perhaps by running all the show commands from a unix script you could dump the output to a file, and at the very least use it as a reference? I imagine it would take considerably more work to convert that text into a format that can be easily sent back to the netserver, but you can always use the windows netserver software if you just need save/restore funtionality. - lv On Thu, 8 Jan 1998, Michael Wronski wrote: > At 12:59 PM 1/8/98 -0800, you wrote: > > > >Can you write the netservers settings out to a text file? > >If so how is that done? > >I posted this question a few days ago and don't know > >if it got lost or what, so please excuss me if someone has > >already addressed this > > > > At this time there is no way to write the config out into a > text file. > > -M > > `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' > 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. >

On Thu, 8 Jan 1998 pat@coffey.com wrote: > I have a new chassis with tcs2.5.1 on it. The card for the span lines > is flashed with the pri code right now. My question is can I flash > this card with the t1 code and have it work? > > Thanks, > Pat Yes, but you'll need to manually type the name of the sdl files in TCM/software dl..if it thinks it's a PRI card, it won't look for T1 code of any vintage. ========================================================================= 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

I have been asked to "help" another company get their netserver to authenticate using the USR/3com RADIUS server against the local password file on a Unix system. There is 1 big catch. I'm 1000 miles from the console and can't get x to display back at this moment. I got them to tak to each other. I can have a user appear to authenticate, and can use an interactive terminal window to log into the netserver using the password from the unix server. This was accomplished by moving my db to the postgres db and using the postgres sql calls (and sautil_pg -v -d) to really look at the database and determine what is munged. BUT.... I can't get windows 95 users to login. It will work if we setup the user on the netserver card. I've been through 3 days of 3com support (different very long rant) and gotten nothing! I think I am looking for either 1) the proper "set" command on the Netserver which I don't know and/or 2) a set of sample user and system config files which are functional. This has never been a problem on other tcm systems with which I have worked. And I am stumped. Tech support has been useless. doug

On Fri, 9 Jan 1998, Eric Forcey wrote: > > I have asked this before and never received a response. > > We have a TC rack with a dual T1 and a dual pri card in it. > > The Pri I am using for ISDN, we have Rip turned off on the Pri card. > > All of my ISDN customers receive subnetted class C's from us, so we add in > static routes to the Netserver card. > > I have one particular customer that is getting 2 subnets from us (due to > their internal network they need 2 separate subnets). I add the routes > into the static table, and every so often (has been more often than not > lately) the second route drops from the table. > > I have called USR and opened tickets with them, but so far they have not > been any help. > > The latest thing that was suggested was turn proxyarp on, which I did, > then also have the netserver listen for routes, which I did as well. It is > still happening. The route dropped again this morning. You are not using any routing protocol - For this senario the only routing protocol that you can use with the NETServer is RIP ver 2. Currently you are not allowing the NETServer to either broadcast or listen to the route. There is no problem with your static route, the the user connects the static route will be up and have a proper metric and when he disconnects the route will have a metric of 16. Think about the problem - you have user who is using a part of your C class network and now you want another part of C class ( the same or different ) to be routed to him. The NETServer is told not to route, so it does not advertise the route. The user may work when he has some staic entry on his side but for the most part will not work. The NETServer will send the packet from the subnetted subnet and put it on the wire - now why should anyone respond to it? - The routing information has to be propogated before someone knows how to respond. You have two choice a. use RIP ver 2 on your network and broadcast/listen b. Use static route - update your router tables krish > > > Anyone ever come across this yet? > > -Eric > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Thus spake Douglas O'Flaherty >I've been through 3 days of 3com support (different very long rant) and >gotten nothing! 3 days? You're just getting started, it took me 9 days to get anyone halfway clueful there (Vito, he's actually pretty sharp), also couldn't get a callback to save my life, despite calling 4 times myself. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

I've made my own text files of every command that it takes to start up a netserver. I telnet in and paste it to the buffer and it types it in for me. Granted, I only have 4 netservers but every time I upgrade or have problems I'm the kind of person who likes to start fresh so I nuke the config and just paste the new one back in. It took about 4 hours but it was worth it. Tom > -----Original Message----- > From: Laszlo Vecsey [SMTP:master@internexus.net] > Sent: Thursday, January 08, 1998 10:21 PM > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) Config files for Netservers. > > hmm, perhaps by running all the show commands from a unix script you > could > dump the output to a file, and at the very least use it as a > reference? > > I imagine it would take considerably more work to convert that text > into a > format that can be easily sent back to the netserver, but you can > always > use the windows netserver software if you just need save/restore > funtionality. > > - lv > > On Thu, 8 Jan 1998, Michael Wronski wrote: > > > At 12:59 PM 1/8/98 -0800, you wrote: > > > > > >Can you write the netservers settings out to a text file? > > >If so how is that done? > > >I posted this question a few days ago and don't know > > >if it got lost or what, so please excuss me if someone has > > >already addressed this > > > > > > > At this time there is no way to write the config out into a > > text file. > > > > -M > > > > > `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' > > 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. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message.

Robert von Bismarck said once upon a time: >Has anyone a list of commands issued to a HiPerARC to make all modems >active, and accept PPP calls (analog and ISDN) authenticated by RADIUS, >getting their IP's from a predefined pool ?? > >Thanks for any info, because I've got a deadline to get a site into >production with over 500 ports loaded into two racks with HiPerDSP's and >HiPerARC's (I hate those marketing guys and their deadlines, don't you Here is what I've stuffed into mine, substitute X.X.X.X for IP addresses and SECRET for passwords: add user guest password "" login_service rlogin type login set login user guest login_host_ip_address X.X.X.X set login user guest terminal_type dialup enable user guest add dns server X.X.X.X preference 2 set dns domain_name xmission.com set modem_group all message "Welcome to XMission Internet Access\r\n\r\n Type 'guest' for Guest Services\r\n\r\n 'account' for PPP\r\n 'account@slip' for SLIP\r\n 'account@shell' for Shell\r\n\r\n" set modem_group all prompt "slc10-tc login: " disable ip network ip set ip network ip rip_policies_update no_ripv1_receive set ip network ip routing_protocol ripv2 set ip network ip rip_policies_update no_ripv1_receive set ip network ip rip_policies_update no_send_compat enable ip network ip set authentication primary_server X.X.X.X set authentication primary_secret SECRET set accounting primary_secret SECRET set system name XMission set system contact support@xmission.com set syslog X.X.X.X facility log_local3 set ppp receive_authentication pap enable ip security_option disallow_source_route_options set ntp primary_server X.X.X.X set ntp secondary_server X.X.X.X Here are my current HDM settings that differ from the default: Carrier loss detect time: 50 Dial length: 50 DTE NVRAM lock: enable Echo DTE: enable Result code groups: 7 ARQ result codes: includeProto response to +++: ignore ANI: 10 DNIS: 7 ISDN v120: enable (this doesn't stick, and has been reported as a bug) Most of these HDM settings apply to outgoing calls. I'd say the "carrier loss," "response to +++," "ANI," "DNIS," and "v120" are important for dialin. Make sure your DNIS matches what the telco is sending you if you are using DSS. Hope that helps.

> > I've made my own text files of every command that it takes to start up a > > netserver. I telnet in and paste it to the buffer and it types it in > for me. > Granted, I only have 4 netservers but every time I upgrade or have > problems > I'm the kind of person who likes to start fresh so I nuke the config and > just paste the new one back in. > > It took about 4 hours but it was worth it. > > Tom Tom: As I am dealing with a netserver without a manual can I ask you to email me that config so I can check it against the way USR setup the netserver for us. It doesn't do PAP correctly, still. doug

Is there a command line on the netserver to show all current set configs? doug

I have asked this before and never received a response. We have a TC rack with a dual T1 and a dual pri card in it. The Pri I am using for ISDN, we have Rip turned off on the Pri card. All of my ISDN customers receive subnetted class C's from us, so we add in static routes to the Netserver card. I have one particular customer that is getting 2 subnets from us (due to their internal network they need 2 separate subnets). I add the routes into the static table, and every so often (has been more often than not lately) the second route drops from the table. I have called USR and opened tickets with them, but so far they have not been any help. The latest thing that was suggested was turn proxyarp on, which I did, then also have the netserver listen for routes, which I did as well. It is still happening. The route dropped again this morning. Anyone ever come across this yet? -Eric

Robert von Bismarck said once upon a time: > >Thanks a lot pal !! > >It works fine with analog, I'll test it with ISDN lines on Monday. This >really helped me out, now I've got a week-end to read the PDF's ;-) One further suggestion on the PDF for the ARC. Go to the local printshop and have them bind you a manual based on the PDF. It makes configuring and working with the ARC a lot easier. Now if only the index had the right page numbers... ;-)

System Administrator said once upon a time: >> I have one particular customer that is getting 2 subnets from us (due to >> their internal network they need 2 separate subnets). I add the routes >> into the static table, and every so often (has been more often than not >> lately) the second route drops from the table. >> >> The latest thing that was suggested was turn proxyarp on, which I did, >> then also have the netserver listen for routes, which I did as well. It is >> still happening. The route dropped again this morning. I would suggest using the RADIUS "Framed-Route" which works fine for multiple routes. You do need 3.5 or higher on your Netserver for non /32's and /24's to propogate with the correct subnet mask under RIPv2. Static routes have worked OK for us in the past, but Framed-Route is a lot easier to manage. We do have a couple customers with two networks routed to them with this method and it has been good.

Douglas O'Flaherty said once upon a time: >Thanks to those who made suggestions, my problem is becoming clearer: > IP Gateway: 207.150.95.1 Gateway Metric: 1 > Assigned Address: 207.150.95.11 Reported Address: 207.150.95.4 >What Have I missed? It looks like you may not be subnetting properly. What are the subnet sizes for all the above addresses?

The name of the dialup daemon for mlppp on FreeBSD is mpd. Has anyone used it successfully with the usr-tc? I'm still not sure if its a problem with the usr-tc, or mpd implementation of multilinkppp. Uploads are ok but downloads are way too slow. - lv

On Fri, 9 Jan 1998, Eric Forcey wrote: > I have asked this before and never received a response. > > We have a TC rack with a dual T1 and a dual pri card in it. > > The Pri I am using for ISDN, we have Rip turned off on the Pri card. > > All of my ISDN customers receive subnetted class C's from us, so we add in > static routes to the Netserver card. > > I have one particular customer that is getting 2 subnets from us (due to > their internal network they need 2 separate subnets). I add the routes > into the static table, and every so often (has been more often than not > lately) the second route drops from the table. > > I have called USR and opened tickets with them, but so far they have not > been any help. > > The latest thing that was suggested was turn proxyarp on, which I did, > then also have the netserver listen for routes, which I did as well. It is > still happening. The route dropped again this morning. > > > Anyone ever come across this yet? I have never had _ANY_ great success getting static routes to work on the NETServer. Occasionally, I have gotten them to function, but not with any sort of clear methodology (to keep them working). I would recommend avoiding static routes at all costs on the NETServer. It sounds like this might be impossible to do in the case of two subnets. However, in all other cases I use the technique of ripv2 on the NETServer with a cisco listening. I aggregate all /32 routes on the cisco and propogate them via OSPF to the rest of my network. Non-/32's are not-aggregated and simply injected right into the OSPF domain. Works like a charm, however I have never had to deal a customer needing two (or more) different distinct subnets. Perhaps some sort of tunneling (depending on what your customer is using for a router)? Regards, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

Thanks to those who made suggestions, my problem is becoming clearer: I'm fairly confident I do not have a legitimate IP collision. I suspect that I have the netserver trying to hand out the same IP twice. What can I check next: Syslog from the netserver for window login 1 is standard, 2 is interactive Jan 9 13:25:40 ponyexpress-tc1 MODEM: S32: CALL_REF >0x0000a33b< PRI_SLOT >0< TS >31< SPAN >0< B_CH >1< Jan 9 13:25:40 ponyexpress-tc1 acct 0x0000a33b dial: S32 call arrived. Jan 9 13:25:40 ponyexpress-tc1 sent out answer incoming call for S32. Jan 9 13:25:53 ponyexpress-tc1 acct 0x0000a33b dial: S32 answered the phone using handle 36. Jan 9 13:26:07 ponyexpress-tc1 NetS: port S32 session disconnected user global Jan 9 13:26:07 ponyexpress-tc1 Debug (PROC_FORCED_DISCONNECT): S32 has non-zero slif address (207.150.95.32) Jan 9 13:26:08 ponyexpress-tc1 acct 0x0000a33b dial: S32 hung up the phone. Call duration 0:0:27. Jan 9 13:26:25 ponyexpress-tc1 MODEM: S33: CALL_REF >0x0000a33c< PRI_SLOT >0< TS >32< SPAN >0< B_CH >1< Jan 9 13:26:25 ponyexpress-tc1 acct 0x0000a33c dial: S33 call arrived. Jan 9 13:26:25 ponyexpress-tc1 sent out answer incoming call for S33. Jan 9 13:26:37 ponyexpress-tc1 acct 0x0000a33c dial: S33 answered the phone using handle 37. Jan 9 13:26:49 ponyexpress-tc1 dialnet: port S33 global login failed Jan 9 13:26:52 ponyexpress-tc1 acct dialnet: port S33 PPP succeeded dest Negotiated Jan 9 13:27:11 ponyexpress-tc1 dialnet: port S33 ppp_sync failed dest Jan 9 13:27:12 ponyexpress-tc1 acct 0x0000a33c dial: S33 hung up the phone. Call duration 0:0:47. config from the netserver: IP Gateway: 207.150.95.1 Gateway Metric: 1 IPX Gateway: 00000000:000000000000 Gateway Metric: 0 Default Route: Broadcast, Listen (On) Telnet Access Port: 23 TCP Maximal Segment Size: 0 Assigned Address: 207.150.95.11 Reported Address: 207.150.95.4 Assigned Pool Size: 128 Periodic CHAP timeout: 5 PPP in modem: ON SLIP in modem: OFF Packet bus clock: SLAVE ICMP error logging: OFF Connect message: OFF Dial !root access: ON Random hosts list: OFF SNMP: ON Proxy Arp: ON Response message: ON PPP message: OFF Lan/Wan Routing: ON RIP V2 Authen: OFF VPN Local Routing: ON MPIP Server: OFF Hint assigned: OFF Tap Login: OFF Syslog facility: local5 Extd. IPXCP Opts: OFF Acct AuthChk: OFF Send DNS info: ON DNS cache reset timeout: 0 days 0 hours 30 minutes (30 min) What Have I missed? doug

Yeah, all this talk about netserver cards is cool, but I just got 12 (!) HiPerARC's, and I stuffed one in my test chassis which has currently two PRI's connected to two HiPerDSP's to check out how they are configured. I configured it with the CLI Wizard, and began reading the PDF. I know some of you guys have some working HiPerARC's, mine will pick up the line, connect say some blabla and prompt me for login and password. My admin username and pass work perfectly well, I can manage my chassis via modem, but I can't get a RADIUS login. It always responds : login incorrect and makes no RADIUS request at all. The server is configured to accept requests from this chassis. Has anyone a list of commands issued to a HiPerARC to make all modems active, and accept PPP calls (analog and ISDN) authenticated by RADIUS, getting their IP's from a predefined pool ?? Thanks for any info, because I've got a deadline to get a site into production with over 500 ports loaded into two racks with HiPerDSP's and HiPerARC's (I hate those marketing guys and their deadlines, don't you ?) Robert von Bismarck Petrel Communication S.A.

Mmm.. do what I do, don't waste time and effort going through the idiots at tech support. Call your sales person and tell them to 'Fix it'. Be firm yet pleasant, and don't let them weasel out by telling you to call the blackhole^H^H^H^tech support line. 3Com, Isn't it about time someone down there in the depths of hell let us sell a piece of our soul for a 2nd level support line so we can avoid the trained monkies answering your regular lines? On Fri, Jan 09, 1998 at 08:15:51AM -0500, Jeff Mcadams made some electrons appear in the following form: > Thus spake Douglas O'Flaherty > >I've been through 3 days of 3com support (different very long rant) and > >gotten nothing! > > 3 days? You're just getting started, it took me 9 days to get anyone > halfway clueful there (Vito, he's actually pretty sharp), also couldn't > get a callback to save my life, despite calling 4 times myself. > -- > 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. --

Greetings to the list.... We are having a perplexing problem. Seems users are getting bumped from our MP/I-16 randomly. Also, sometimes when a user hangs up normally, another user who did not wish to hang up gets dumped. This acts like it's a line noise problem as near as I can tell, and so I've set the S7 value to 20 from 7 but this did not seem to have any effect on the problem. I can find no reason for this to be happening in the logs, nor can I make it happen on demand in any way. Does anyone have any ideas about what could be causing this? Len Haggblad Mainline Information Service MtnMail V3.05: A bird in the hand...tickles.

Where can one get the latest TCM Software besides the totalservice web site? I have registered on the web site and am able to successfully login. However, clicking on files says that they are still locked. FTP to totalservice.usr.com, will not allow my login and password to work either. How did everyone else on this site get their login/passwords? I'm not paying an arm and leg for a support contract when all I need is one program that should be included with the unit. Lee Kuo

On Fri, 9 Jan 1998 lhaggblad@mainline.ab.ca wrote: > We are having a perplexing problem. Seems users are getting bumped from > our MP/I-16 randomly. Also, sometimes when a user hangs up normally, > another user who did not wish to hang up gets dumped. This acts like > it's a line noise problem as near as I can tell, and so I've set the S7 > value to 20 from 7 but this did not seem to have any effect on the > problem. > > I can find no reason for this to be happening in the logs, nor can I > make it happen on demand in any way. Does anyone have any ideas about > what could be causing this? After a couple of months with between 10 and 20 multilink connections on a Netserver card running 3.4.23 we would start to see customers getting disconnected when others dropped and users logging on and getting the IP address of the customer whose B-channel had just dropped from that I port. We had users with up to 4 B-channels connected intermittently (on-demand) for a while and this accellerated the corruption process. We just had to reboot it. Newer versions may not have this problem. -CH

Thanks a lot pal !! It works fine with analog, I'll test it with ISDN lines on Monday. This really helped me out, now I've got a week-end to read the PDF's ;-) Robert -----Original Message----- From: Pete Ashdown [SMTP:pashdown@xmission.com] Sent: vendredi, 9. janvier 1998 16:53 To: usr-tc@lists.xmission.com Subject: Re: (usr-tc) Config files for HiPerARCs Robert von Bismarck said once upon a time: >Has anyone a list of commands issued to a HiPerARC to make all modems >active, and accept PPP calls (analog and ISDN) authenticated by RADIUS, >getting their IP's from a predefined pool ?? > >Thanks for any info, because I've got a deadline to get a site into >production with over 500 ports loaded into two racks with HiPerDSP's and >HiPerARC's (I hate those marketing guys and their deadlines, don't you Here is what I've stuffed into mine, substitute X.X.X.X for IP addresses and SECRET for passwords: add user guest password "" login_service rlogin type login set login user guest login_host_ip_address X.X.X.X set login user guest terminal_type dialup enable user guest add dns server X.X.X.X preference 2 set dns domain_name xmission.com set modem_group all message "Welcome to XMission Internet Access\r\n\r\n Type 'guest' for Guest Services\r\n\r\n 'account' for PPP\r\n 'account@slip' for SLIP\r\n 'account@shell' for Shell\r\n\r\n" set modem_group all prompt "slc10-tc login: " disable ip network ip set ip network ip rip_policies_update no_ripv1_receive set ip network ip routing_protocol ripv2 set ip network ip rip_policies_update no_ripv1_receive set ip network ip rip_policies_update no_send_compat enable ip network ip set authentication primary_server X.X.X.X set authentication primary_secret SECRET set accounting primary_secret SECRET set system name XMission set system contact support@xmission.com set syslog X.X.X.X facility log_local3 set ppp receive_authentication pap enable ip security_option disallow_source_route_options set ntp primary_server X.X.X.X set ntp secondary_server X.X.X.X Here are my current HDM settings that differ from the default: Carrier loss detect time: 50 Dial length: 50 DTE NVRAM lock: enable Echo DTE: enable Result code groups: 7 ARQ result codes: includeProto response to +++: ignore ANI: 10 DNIS: 7 ISDN v120: enable (this doesn't stick, and has been reported as a bug) Most of these HDM settings apply to outgoing calls. I'd say the "carrier loss," "response to +++," "ANI," "DNIS," and "v120" are important for dialin. Make sure your DNIS matches what the telco is sending you if you are using DSS. Hope that helps. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.

trained monkies???? do you really thing they train the main line tech support people? i really doubt it. the folks i have had are either level II techs who got stuck on the phone (rarely) or they didn't even understand what i was talking about. not just not understanding the problem but not even seeming to understand the product. matthew Brett Hawn wrote: > > Mmm.. do what I do, don't waste time and effort going through the idiots at > tech support. Call your sales person and tell them to 'Fix it'. Be firm yet > pleasant, and don't let them weasel out by telling you to call the > blackhole^H^H^H^tech support line. > > 3Com, > > Isn't it about time someone down there in the depths of hell let us sell a > piece of our soul for a 2nd level support line so we can avoid the trained > monkies answering your regular lines? > > On Fri, Jan 09, 1998 at 08:15:51AM -0500, Jeff Mcadams made some electrons appear in the following form: > > Thus spake Douglas O'Flaherty > > >I've been through 3 days of 3com support (different very long rant) and > > >gotten nothing! > > > > 3 days? You're just getting started, it took me 9 days to get anyone > > halfway clueful there (Vito, he's actually pretty sharp), also couldn't > > get a callback to save my life, despite calling 4 times myself. > > -- > > 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. -- the spa! online services - www.the-spa.com - info@the-spa.com 654b new ludlow road, south hadley, ma 01075 - 413 539-9818 providing online services since 1989!

At 03:45 PM 1/9/98 -0600, Brett Hawn wrote: >3Com, > > Isn't it about time someone down there in the depths of hell let us sell a >piece of our soul for a 2nd level support line so we can avoid the trained >monkies answering your regular lines? > I understand that our SA (Brian) can now skip level 1 and go straight to level 2. How he aquired this privledge I'm not quite sure. Either by the grace of God or he took a test or something. But this really is positive for both parties.. #1 It will shave days off the time required to actually start talking about the real problem and working toward a real solution. #2 It will save $10-$20 in unnecessary telco expenses (800) and twice that in labor on our end. I don't mind spending alittle. (and neither does USR on 800 service). But I wish the money would go toward increasing level 2 staff or salaries.. They're the ones who really deserve it.. #3 It will keep Brian from sharing his level 1 "experiences" with the rest of the planet. It seems like every competent TC admin has a level 1 story to tell.. Some are pretty entertaining.. (to the rest of us ;) But it's not funny when it's *your* hub thats down.. And your listening to THAT music.. #4 Other corporations have certification programs, etc.. Although most are pretty lame, the concept is a logical one. And some corporations do a good job (and make a profit I imagine) with education/certification. In any event, if I ran things at USR, it would be a requirement that all level 1 staff READ THIS LIST. And hopefully understand most of it. Our local tech staff all read isp-tech which is a good resource in their field. If you don't read it, it's a wasted resource. I think alot of Krish for what he does on this list. He is good at it and seems to enjoy his work. But aren't there others at USR/3COM that would also benefit from this list. I even think it would be good for management/sales to "lurk" and see what folks think and say about their flagship product. I doubt any do because old "harped upon" problems (like phone support) seem go go on forever.. My .02, IMHO, and all that... 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

> > I think alot of Krish for what he does on this list. He is good at it and > seems to enjoy his work. But aren't there others at USR/3COM that would also > benefit from this list. I even think it would be good for management/sales > to "lurk" and see what folks think and say about their flagship product. > I doubt any do because old "harped upon" problems (like phone support) seem > go go on forever.. My .02, IMHO, and all that... > > Allen > I think reading this list, for 3com tech support would be good, just becuase you learn alot about some of the problems, but then again, as level 1 they learn alot answering calls all day. One thing I would require of level 1, is that they become intimatly familiar with the Users manuals for the enterprise hardware they do tech support for. Especially the troubleshooting procedures. The second thing I would require is that they know about the "open" issues with hardware. These two things are easy for them to accomplish, and would give them good merit. I have seen some REALLY sharp 3com level 1 support techs, such as Toan (pronounced Twon, who is probably level 2 at least by now). I feel there isn't enough people like him around or else 3com WOULD hire more, they have to take what they can get, and train as best they can. Skip Lasher(sp?) who is a Supervisor for 3com tech support, did talk to me a while back, and he addressed some of the tech support issues. 3com is aware that there tech support could be better trained, and he assured me they are working on it. They were trying to concentrate on getting hold queue times down, and now that they have done a good job at that (you have to give them credit, it has gotten better), they are working on getting there people up to speed. Also, level 1 may not be that great, but are they in ANY company? I mean, they ARE level 1, they arent suppose to be wizzes, but I agree, there are certain things they should know. 3com Level II, is very good. Ken Tam(sp?) and some of those other guys really know there stuff, and can be very valuable. They also know the open issues/field reports and are very resourcefull. A level of competency such as some of of you all have from ISP's running many hubs with lots of experience (xmission, texas.net, etc) shouldn't have to talk to level 1, and I do beleive there is a test you can take to afford you better support, and no doubt you all would pass it. Brian > > > _____________________________________________________________ > 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 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

Over the past YEAR, I have had ONE customer (that I am aware of) experience the problem below. Now I have another one. My "gut instinct" tells me that this is not related to usr-tc, however I thought I would ask here anyway because the clue factor seems to be quite high on this list. :-) A customer begins an FTP transfer from from a local server on our network to their home win95 system. In this case the customer has a solid x2 connection. Initially they experience a throughput of approx. 5KB/s (reasonable for x2), however the throughput continuously degrades over the period of the transfer until they are getting an average of 1KB/s. I have had the customer install the lastest MS DUN, with no positive results. I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN I get a consistant 14KB/s. Over a test x2 connection I am able to sustain throughput of 4-5KB/s using the SAME FTP client that the customer is using, transfering the same file from the same FTP server. The customer's modem is internal, and we have fiddled with every setting imaginable to man (baud rates, ip header compression, etc, etc), with no success. As I said before, the almost EXACT same thing happened with another customer in 1997. Anyone ever seen this before, or have any suggestions? Regards, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

System Administrator said once upon a time: >connection. Initially they experience a throughput of approx. 5KB/s >(reasonable for x2), however the throughput continuously degrades over the >period of the transfer until they are getting an average of 1KB/s. I have >had the customer install the lastest MS DUN, with no positive results. > >I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN ISDN is not X2. My guess is that you may be seeing a problem with X2, possibly due to retrains. I know very little about actual analog transmissions, but I would presume that sending a lot of data over one (as in a FTP transfer) would stress it more than without. Lets hope X2v2 (whenever it comes) will fix a lot of these idiosyncrasies.

> Date: Fri, 9 Jan 1998 23:45:48 -0600 (CST) > From: Brian <signal@shreve.net> > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Subject: Re: (usr-tc) Tech-Support rant > Reply-to: usr-tc@lists.xmission.com > > > > I think alot of Krish for what he does on this list. He is good at it and > > seems to enjoy his work. But aren't there others at USR/3COM that would also > > benefit from this list. I even think it would be good for management/sales > > to "lurk" and see what folks think and say about their flagship product. > > I doubt any do because old "harped upon" problems (like phone support) seem > > go go on forever.. My .02, IMHO, and all that... > > > > Allen > > > > I think reading this list, for 3com tech support would be good, just > becuase you learn alot about some of the problems, but then again, as > level 1 they learn alot answering calls all day. > > One thing I would require of level 1, is that they become intimatly > familiar with the Users manuals for the enterprise hardware they do tech > support for. Especially the troubleshooting procedures. > > The second thing I would require is that they know about the "open" issues > with hardware. > > These two things are easy for them to accomplish, and would give them good > merit. I have seen some REALLY sharp 3com level 1 support techs, such as > Toan (pronounced Twon, who is probably level 2 at least by now). I feel > there isn't enough people like him around or else 3com WOULD hire more, > they have to take what they can get, and train as best they can. > > Skip Lasher(sp?) who is a Supervisor for 3com tech support, did talk to me > a while back, and he addressed some of the tech support issues. 3com is > aware that there tech support could be better trained, and he assured me > they are working on it. They were trying to concentrate on getting hold > queue times down, and now that they have done a good job at that (you have Yes, I have heard this mumbo jumbo from them as well....Come on, lets be realistic....What good is getting the hold times down from 45 minutes to 5 minutes if the person on the phone that picks up still does not know what he is talking about....Everyone knows at 3com that level on techs are trained and given a book of common problems, if your problem is not in the book there is no answer.....That would be alright if after the level 1's realized this they escalated it to a level 2 like they should be told to do....(Maybe they are not)...This problem is very easy to solve and if I did not have my hands tied to my business 18 hours a day I would apply for the job......I would not mind waiting 45 minutes on hold if I knew my problem was going to get resolved without having to call back 9 times....Also, the thing about that is that everytime you call back you talk to another level one and they all have the notes and your trouble ticket and they see that the other 7 people have not been able to help you but they still do not escalate to level 2 unless you throw a fit and demand it.....This is bad service....Now, I got the support on one of my units for the first 90 days and during the week it is the same service I would get if I purchased the maintenance contract for 2,500 a yeat...They want me to purchase a 2500 a year(or whatever) contract to call 9 times and still not get an answer....The only difference with the contract is that it is call premium support and you also get to call on the weekends if it is important......You get a tech to call back on the weekends and they still dont know whats up.......I am done for now....But I could go one for another hour about this 3com garbage....Thank you..... > to give them credit, it has gotten better), they are working on getting > there people up to speed. > > Also, level 1 may not be that great, but are they in ANY company? I mean, > they ARE level 1, they arent suppose to be wizzes, but I agree, there are > certain things they should know. > > 3com Level II, is very good. Ken Tam(sp?) and some of those other guys > really know there stuff, and can be very valuable. They also know the > open issues/field reports and are very resourcefull. > > A level of competency such as some of of you all have from ISP's running > many hubs with lots of experience (xmission, texas.net, etc) shouldn't > have to talk to level 1, and I do beleive there is a test you can take to > afford you better support, and no doubt you all would pass it. > > Brian > > > > > > > > > _____________________________________________________________ > > 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 > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. > Richard Mazurowski SurfNet Corporation Voice (773) 283-9000 Fax (773) 283-9009

US>> We are having a perplexing problem. Seems users are getting bumped from US>> our MP/I-16 randomly. Also, sometimes when a user hangs up normally, US>> another user who did not wish to hang up gets dumped. This acts like US>> it's a line noise problem as near as I can tell, and so I've set the S7 US>> value to 20 from 7 but this did not seem to have any effect on the US>> problem. US>> US>> I can find no reason for this to be happening in the logs, nor can I US>> make it happen on demand in any way. Does anyone have any ideas about US>> what could be causing this? US>After a couple of months with between 10 and 20 multilink connections on a US>Netserver card running 3.4.23 we would start to see customers getting US>disconnected when others dropped and users logging on and getting the IP US>address of the customer whose B-channel had just dropped from that I port. US>We had users with up to 4 B-channels connected intermittently (on-demand) US>for a while and this accellerated the corruption process. We just had to US>reboot it. Newer versions may not have this problem. Thanks for your reply, but I have tried rebooting many times to no avail. Also please note that this is a MP/I-16 not a Netserver. Len Haggblad Mainline Information Service MtnMail V3.05: Anything that can be changed will be changed until there is no time left

MtnMail V3.05: Anything that can be changed will be changed until there is no time left

MtnMail V3.05: Anything that can be changed will be changed until there is no time left

Just be sure to remember there are TWO different versions of the T1 software. The original version that used the old 186 processors, and the new version that runs on the Dual PRI cards. They are NOT interchangable. At 09:05 PM 1/8/98 -0600, you wrote: >On Thu, 8 Jan 1998 pat@coffey.com wrote: > >> I have a new chassis with tcs2.5.1 on it. The card for the span lines >> is flashed with the pri code right now. My question is can I flash >> this card with the t1 code and have it work? >> >> Thanks, >> Pat >> >> > >yep, just flash it T1. Ours came T1, and we flashed PRI. > > >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the 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. > >

On Sat, 10 Jan 1998, Pete Ashdown wrote: > System Administrator said once upon a time: > > >connection. Initially they experience a throughput of approx. 5KB/s > >(reasonable for x2), however the throughput continuously degrades over the > >period of the transfer until they are getting an average of 1KB/s. I have > >had the customer install the lastest MS DUN, with no positive results. > > > >I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN > > ISDN is not X2. Right. This is why I also did testing w/ an x2 modem (as stated in previous message). Actually, at this point I even had the customer DISABLE x2, and exactly the same thing happens. > My guess is that you may be seeing a problem with X2, > possibly due to retrains. I know very little about actual analog > transmissions, but I would presume that sending a lot of data over one (as > in a FTP transfer) would stress it more than without. Lets hope X2v2 > (whenever it comes) will fix a lot of these idiosyncrasies. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

Is the customer running a 56K Sportster? We have found that the only time there is a problem with spiraling (down) throughput is when a 56K Sportster is connected to the TC Hub. Flashing to the latest X2 code only worsens the problem. There are many freezes that sporadically happen throughout the download the lower the throughput rate. Other non-Sportster modems do not experience this problem. The only solution that we have been able to come up with (after opening a dozen or so trouble tickets with 3COM), is to NOT reboot the hub. It seems after rebooting the hub this problem is very serious and after a couple days, seems to happen less and less. Now, approximately 45 days after the reboot, it only happens once in a while. 3COM support refuses to believe there is a problem and is unable to test it (they only have Couriers, no Sportsters there). On Sat, 10 Jan 1998, System Administrator wrote: > Over the past YEAR, I have had ONE customer (that I am aware of) > experience the problem below. Now I have another one. My "gut instinct" > tells me that this is not related to usr-tc, however I thought I would ask > here anyway because the clue factor seems to be quite high on this list. > :-) > A customer begins an FTP transfer from from a local server on our network > to their home win95 system. In this case the customer has a solid x2 > connection. Initially they experience a throughput of approx. 5KB/s > (reasonable for x2), however the throughput continuously degrades over the > period of the transfer until they are getting an average of 1KB/s. I have > had the customer install the lastest MS DUN, with no positive results. > I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN > I get a consistant 14KB/s. Over a test x2 connection I am able to sustain > throughput of 4-5KB/s using the SAME FTP client that the customer is > using, transfering the same file from the same FTP server. The customer's > modem is internal, and we have fiddled with every setting imaginable to > man (baud rates, ip header compression, etc, etc), with no success. As I > said before, the almost EXACT same thing happened with another customer in > 1997.

I just got my TCM software downloaded (windows version 5.0.4) and I can connect to my TC hub. It will go and discover all the cards. When I go to highlight any card and select CONFIGURE/PROGRAMMED SETTINGS I get the folling error. "Error Opening Device Description File." It does this for every slot in the chassis. Anyhow, I am new to this so forgive me if I am overlooking something, but I can't seem to get past this. Does anyone have a recommendation on what to do? THANKS! ============================================================================= 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. =============================================================================

Sounds like the old classic Sportster Spiraling Death Syndrome. On Sat, 10 Jan 1998, System Administrator wrote: > Over the past YEAR, I have had ONE customer (that I am aware of) > experience the problem below. Now I have another one. My "gut instinct" > tells me that this is not related to usr-tc, however I thought I would ask > here anyway because the clue factor seems to be quite high on this list. > :-) > > A customer begins an FTP transfer from from a local server on our network > to their home win95 system. In this case the customer has a solid x2 > connection. Initially they experience a throughput of approx. 5KB/s > (reasonable for x2), however the throughput continuously degrades over the > period of the transfer until they are getting an average of 1KB/s. I have > had the customer install the lastest MS DUN, with no positive results. > > I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN > I get a consistant 14KB/s. Over a test x2 connection I am able to sustain > throughput of 4-5KB/s using the SAME FTP client that the customer is > using, transfering the same file from the same FTP server. The customer's > modem is internal, and we have fiddled with every setting imaginable to > man (baud rates, ip header compression, etc, etc), with no success. As I > said before, the almost EXACT same thing happened with another customer in > 1997. > > Anyone ever seen this before, or have any suggestions? > > 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. >

Sounds like the old classic Sportster Spiraling Death Syndrome. On Sat, 10 Jan 1998, System Administrator wrote: > Over the past YEAR, I have had ONE customer (that I am aware of) > experience the problem below. Now I have another one. My "gut instinct" > tells me that this is not related to usr-tc, however I thought I would ask > here anyway because the clue factor seems to be quite high on this list. > :-) > > A customer begins an FTP transfer from from a local server on our network > to their home win95 system. In this case the customer has a solid x2 > connection. Initially they experience a throughput of approx. 5KB/s > (reasonable for x2), however the throughput continuously degrades over the > period of the transfer until they are getting an average of 1KB/s. I have > had the customer install the lastest MS DUN, with no positive results. > > I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN > I get a consistant 14KB/s. Over a test x2 connection I am able to sustain > throughput of 4-5KB/s using the SAME FTP client that the customer is > using, transfering the same file from the same FTP server. The customer's > modem is internal, and we have fiddled with every setting imaginable to > man (baud rates, ip header compression, etc, etc), with no success. As I > said before, the almost EXACT same thing happened with another customer in > 1997. > > Anyone ever seen this before, or have any suggestions? > > 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. >

On Sat, 10 Jan 1998, Brian Elfert wrote: > On Sat, 10 Jan 1998, Eric J. Merkel wrote: > > > > > I just got my TCM software downloaded (windows version 5.0.4) and I can > > connect to my TC hub. It will go and discover all the cards. When I > > go to highlight any card and select CONFIGURE/PROGRAMMED SETTINGS I > > get the folling error. "Error Opening Device Description File." > > It does this for every slot in the chassis. > > I'm pretty sure the TCM software has to unzipped with the -D option on > pkunzip. If you unzipped everything into the same directory, the install > might not work right. There is a directory that contains the MIB files > which might be missing in your case. > > Brian > Brian, Thanks for the response. I did unzip with the -d option. There is a folder under the TCM software called MIBS. Anyhow, there are several files in that directory and I downloaded the MIBS off of totalservice and still no luck. Any other ideas? Thanks again! ============================================================================= 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. =============================================================================

On Sat, 10 Jan 1998, Lee Kuo wrote: > Is the customer running a 56K Sportster? We have found that the only time > there is a problem with spiraling (down) throughput is when a 56K > Sportster is connected to the TC Hub. Flashing to the latest X2 code only > worsens the problem. There are many freezes that sporadically happen > throughout the download the lower the throughput rate. Other > non-Sportster modems do not experience this problem. The only solution > that we have been able to come up with (after opening a dozen or so > trouble tickets with 3COM), is to NOT reboot the hub. It seems after > rebooting the hub this problem is very serious and after a couple days, > seems to happen less and less. Now, approximately 45 days after the > reboot, it only happens once in a while. 3COM support refuses to believe > there is a problem and is unable to test it (they only have Couriers, no > Sportsters there). Actually, yes ... the customer does have a Sportster. The hub hasn't been rebooted in quite some time, however I AM running the latest code everywhere. Regards, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

While I have had horror stories of my own, and I know every song on the music-on-hold loop, my last two problems were resolved rather quickly. The last gentleman I spoke with (Brian ??) was excellent, and even pointed me to this list. I guess it's hit and miss, but in general it's gotten a lot better. I've actually had people on the phone who seemed *concerned* that something was broken and were very apologetic for leaving me on hold while they conferred with other techs. Brian, whoever you are, thanks much! C ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Fri, 9 Jan 1998, Brian wrote: > Date: Fri, 9 Jan 1998 23:45:48 -0600 (CST) > From: Brian <signal@shreve.net> > Reply-To: usr-tc@lists.xmission.com > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Subject: Re: (usr-tc) Tech-Support rant > > > > > I think alot of Krish for what he does on this list. He is good at it and > > seems to enjoy his work. But aren't there others at USR/3COM that would also > > benefit from this list. I even think it would be good for management/sales > > to "lurk" and see what folks think and say about their flagship product. > > I doubt any do because old "harped upon" problems (like phone support) seem > > go go on forever.. My .02, IMHO, and all that... > > > > Allen > > > > I think reading this list, for 3com tech support would be good, just > becuase you learn alot about some of the problems, but then again, as > level 1 they learn alot answering calls all day. > > One thing I would require of level 1, is that they become intimatly > familiar with the Users manuals for the enterprise hardware they do tech > support for. Especially the troubleshooting procedures. > > The second thing I would require is that they know about the "open" issues > with hardware. > > These two things are easy for them to accomplish, and would give them good > merit. I have seen some REALLY sharp 3com level 1 support techs, such as > Toan (pronounced Twon, who is probably level 2 at least by now). I feel > there isn't enough people like him around or else 3com WOULD hire more, > they have to take what they can get, and train as best they can. > > Skip Lasher(sp?) who is a Supervisor for 3com tech support, did talk to me > a while back, and he addressed some of the tech support issues. 3com is > aware that there tech support could be better trained, and he assured me > they are working on it. They were trying to concentrate on getting hold > queue times down, and now that they have done a good job at that (you have > to give them credit, it has gotten better), they are working on getting > there people up to speed. > > Also, level 1 may not be that great, but are they in ANY company? I mean, > they ARE level 1, they arent suppose to be wizzes, but I agree, there are > certain things they should know. > > 3com Level II, is very good. Ken Tam(sp?) and some of those other guys > really know there stuff, and can be very valuable. They also know the > open issues/field reports and are very resourcefull. > > A level of competency such as some of of you all have from ISP's running > many hubs with lots of experience (xmission, texas.net, etc) shouldn't > have to talk to level 1, and I do beleive there is a test you can take to > afford you better support, and no doubt you all would pass it. > > Brian > > > > > > > > > _____________________________________________________________ > > 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 > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. >

While I have had horror stories of my own, and I know every song on the music-on-hold loop, my last two problems were resolved rather quickly. The last gentleman I spoke with (Brian ??) was excellent, and even pointed me to this list. I guess it's hit and miss, but in general it's gotten a lot better. I've actually had people on the phone who seemed *concerned* that something was broken and were very apologetic for leaving me on hold while they conferred with other techs. Brian, whoever you are, thanks much! C ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Fri, 9 Jan 1998, Brian wrote: > Date: Fri, 9 Jan 1998 23:45:48 -0600 (CST) > From: Brian <signal@shreve.net> > Reply-To: usr-tc@lists.xmission.com > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Subject: Re: (usr-tc) Tech-Support rant > > > > > I think alot of Krish for what he does on this list. He is good at it and > > seems to enjoy his work. But aren't there others at USR/3COM that would also > > benefit from this list. I even think it would be good for management/sales > > to "lurk" and see what folks think and say about their flagship product. > > I doubt any do because old "harped upon" problems (like phone support) seem > > go go on forever.. My .02, IMHO, and all that... > > > > Allen > > > > I think reading this list, for 3com tech support would be good, just > becuase you learn alot about some of the problems, but then again, as > level 1 they learn alot answering calls all day. > > One thing I would require of level 1, is that they become intimatly > familiar with the Users manuals for the enterprise hardware they do tech > support for. Especially the troubleshooting procedures. > > The second thing I would require is that they know about the "open" issues > with hardware. > > These two things are easy for them to accomplish, and would give them good > merit. I have seen some REALLY sharp 3com level 1 support techs, such as > Toan (pronounced Twon, who is probably level 2 at least by now). I feel > there isn't enough people like him around or else 3com WOULD hire more, > they have to take what they can get, and train as best they can. > > Skip Lasher(sp?) who is a Supervisor for 3com tech support, did talk to me > a while back, and he addressed some of the tech support issues. 3com is > aware that there tech support could be better trained, and he assured me > they are working on it. They were trying to concentrate on getting hold > queue times down, and now that they have done a good job at that (you have > to give them credit, it has gotten better), they are working on getting > there people up to speed. > > Also, level 1 may not be that great, but are they in ANY company? I mean, > they ARE level 1, they arent suppose to be wizzes, but I agree, there are > certain things they should know. > > 3com Level II, is very good. Ken Tam(sp?) and some of those other guys > really know there stuff, and can be very valuable. They also know the > open issues/field reports and are very resourcefull. > > A level of competency such as some of of you all have from ISP's running > many hubs with lots of experience (xmission, texas.net, etc) shouldn't > have to talk to level 1, and I do beleive there is a test you can take to > afford you better support, and no doubt you all would pass it. > > Brian > > > > > > > > > _____________________________________________________________ > > 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 > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. >

> I just got my TCM software downloaded (windows version 5.0.4) and I can > connect to my TC hub. It will go and discover all the cards. When I > go to highlight any card and select CONFIGURE/PROGRAMMED SETTINGS I > get the folling error. "Error Opening Device Description File." > It does this for every slot in the chassis. > > Anyhow, I am new to this so forgive me if I am overlooking something, > but I can't seem to get past this. Does anyone have a recommendation > on what to do? Hmm, this happens to me if NMC card has firmware version that TCM doesn't know about. If NMC reports compatibility version to 5.0.0, TCM searches for MIB files in directory TCM\TCM_DAT\NM050000\ - if it cannot find this it gives that error message. > THANKS! > Eric Merkel | URL: www.metalink.net | Local Access in Kamil Kukura Czech Republic

At 02:26 PM 1/12/98 +0100, you wrote: >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). The list_active_interfaces doesn't work, it always says n+4 >with n being the maximum amount supported by the system. > >Thank you, > Hi, I do that here locally, except I never was able to do it via SNMP. Instead I do it with pmwho. If you have pmwho let me know & I can send you the mrtg.cfg file & another small script you need. Take Care, Elio M. Fuentes Agetech.Net emf@agetech.net

Hi, I've got some stuff that I've hacked together to work with mrtg. The script here was originally something to monitor portmasters, but I changed it around a bit. You will need the ucd-snmp package to get the "snmpwalk" command used in the script... ## CUT HERE ## #!/bin/csh # # tchgather - get a count of ports in use for mrtg # run this out of cron with mrtg # set the names or IP's of your TCHs set TCH1="xxx.xxx.xxx.3" set TCH2="xxx.xxx.xxx.5" set TCH3="xxx.xxx.xxx.7" set TCH4="xxx.xxx.xxx.9" set TCH5="xxx.xxx.xxx.11" # set the SNMP read community of your TCHs set TCH1comm="snmp-read" set TCH2comm="snmp-read" set TCH3comm="snmp-read" set TCH4comm="snmp-read" set TCH5comm="snmp-read" # Grab a user count for each TCH via the snmpwalk command # You may need to use "snmpwalk -v 1" depending on your version of snmpwalk # You may also need to grep for something other than "23" set TC1=`/usr/local/bin/snmpwalk -v 1 $TCH1 $TCH1comm interfaces.ifTable.ifEntry.ifType|grep 23 |wc -l` set TC2=`/usr/local/bin/snmpwalk -v 1 $TCH2 $TCH2comm interfaces.ifTable.ifEntry.ifType|grep 23 |wc -l` set TC3=`/usr/local/bin/snmpwalk -v 1 $TCH3 $TCH3comm interfaces.ifTable.ifEntry.ifType|grep 23 |wc -l` set TC4=`/usr/local/bin/snmpwalk -v 1 $TCH4 $TCH4comm interfaces.ifTable.ifEntry.ifType|grep 23 |wc -l` set TC5=`/usr/local/bin/snmpwalk -v 1 $TCH5 $TCH5comm interfaces.ifTable.ifEntry.ifType|grep 23 |wc -l` # Get a grand total of users online @ TOTAL=$TC1 + $TC2 + $TC3 + $TC4 + $TC5 # MRTG needs 4 lines of input to be happy, we're only interested in the 2nd # line and possibly the first so we fill the others with junk # write out the various logs # grand total of all your TCHs echo "0" > /path/to/mrtgstuff/tchgrand echo $TOTAL >> /path/to/mrtgstuff/tchgrand echo "0" >> /path/to/mrtgstuff/tchgrand echo "0" >> /path/to/mrtgstuff/tchgrand # TCH #1 echo "0" > /path/to/mrtgstuff/tch1-log echo $TC1 >> /path/to/mrtgstuff/tch1-log echo "time" >> /path/to/mrtgstuff/tch1-log echo "tch1" >> /path/to/mrtgstuff/tch1-log # TCH #2 echo "0" > /path/to/mrtgstuff/tch2-log echo $TC2 >> /path/to/mrtgstuff/tch2-log echo "time" >> /path/to/mrtgstuff/tch2-log echo "tch2" >> /path/to/mrtgstuff/tch2-log # TCH #3 echo "0" > /path/to/mrtgstuff/tch3-log echo $TC3 >> /path/to/mrtgstuff/tch3-log echo "time" >> /path/to/mrtgstuff/tch3-log echo "tch3" >> /path/to/mrtgstuff/tch3-log # TCH #4 echo "0" > /path/to/mrtgstuff/tch4-log echo $TC4 >> /path/to/mrtgstuff/tch4-log echo "time" >> /path/to/mrtgstuff/tch4-log echo "tch4" >> /path/to/mrtgstuff/tch4-log # TCH 5# echo "0" > /path/to/mrtgstuff/tch5-log echo $TC5 >> /path/to/mrtgstuff/tch5-log echo "time" >> /path/to/mrtgstuff/tch5-log echo "tch5" >> /path/to/mrtgstuff/tch5-log ## CUT HERE ## And the mrtg config looks like this: # # Experiment - check number of ports active on USRs # Target[USR]: `cat /path/to/mrtgstuff/tchgrand` Title[USR]: Ports active at your location PageTop[USR]: <center><H1>Ports in use at your location</H1></CENTER> # Max bytes should be set to the # of ports you have MaxBytes[USR]: 240 AbsMax[USR]: 240 Options[USR]: absolute, gauge #ShortLegend[USR]: Ports in Use (doesn't seem to work) Unscaled[USR]: dmwy #LegendO[USR]: Ports in Use (doesn't seem to work) Hope this helps, let me know if there are any problems... Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Mon, 12 Jan 1998, Robert von Bismarck wrote: > Date: Mon, 12 Jan 1998 14:26:52 +0100 > From: Robert von Bismarck <rvb@petrel.ch> > Reply-To: usr-tc@lists.xmission.com > To: "'usr-tc@xmission.com'" <usr-tc@xmission.com> > Subject: (usr-tc) SNMP question > > 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). The list_active_interfaces doesn't work, it always says n+4 > with n being the maximum amount supported by the system. > > Thank you, > > Robert von Bismarck > Petrel Communication S.A. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

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). The list_active_interfaces doesn't work, it always says n+4 with n being the maximum amount supported by the system. Thank you, Robert von Bismarck Petrel Communication S.A.

At 03:06 PM 1/12/98 -0500, you wrote: >What is "pmwho"? > Freeware C program written for Liv PM's that telnets into the box and gets connected user information.. Developed since neither the PM or the Netserver can supply that kind of information via SNMP.. Many have modified it to do thier own bidding as well.. There is also a perl program out there that does the same thing.. -M `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)

I've had quite a few "complaints" that after installing our Enterprise Hubs, there is a noticably longer delay during the Windows 95 "Verifying username/password" sequence. I've check the radius accounting logs for possible symptoms, like increased Acct-Delay for trouble communicating with the radius server (auth and acct on same machine). Any suggestions? ========================================================================= 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

At 02:28 PM 1/12/98 -0600, you wrote: >I've had quite a few "complaints" that after installing our >Enterprise Hubs, there is a noticably longer delay during >the Windows 95 "Verifying username/password" sequence. I've >check the radius accounting logs for possible symptoms, >like increased Acct-Delay for trouble communicating with the >radius server (auth and acct on same machine). > >Any suggestions? > Make sure your clients dont have "log into network" checked and that they are only configured for the protocols that you support... Usually unchecking IPX and NETBEUI for the typical ISP setup. This will reduce the LCP negotiation time.. -M

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 \ \-----------------------------------------------------------------------/

What is "pmwho"? ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Mon, 12 Jan 1998, Elio M. Fuentes wrote: > Date: Mon, 12 Jan 1998 10:10:17 -0500 > From: "Elio M. Fuentes" <emf@agetech.net> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) SNMP question > > At 02:26 PM 1/12/98 +0100, you wrote: > >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). The list_active_interfaces doesn't work, it always says n+4 > >with n being the maximum amount supported by the system. > > > >Thank you, > > > > Hi, > > I do that here locally, except I never was able to do it via SNMP. > Instead I do it with pmwho. If you have pmwho let me know & I can send you > the mrtg.cfg file & another small script you need. > > Take Care, > > Elio M. Fuentes > Agetech.Net > > emf@agetech.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. >

On Mon, 12 Jan 1998, Michael Wronski wrote: Is there a bug in Netserver 16I dynamic IP address assignment that causes it to assign address beyond the range of Start_address + Num_ports For example: Netserver 16I assigned address: x.x.x.202 Assigns addresses up to: x.x.x.218 Heres a grep | sort -u on IP pool addresses assigned. This 16 port machine thinks it's IP address pool size is 17! Maybe they're counting the console port? Framed-IP-Address = 209.100.92.202 Framed-IP-Address = 209.100.92.203 Framed-IP-Address = 209.100.92.204 Framed-IP-Address = 209.100.92.205 Framed-IP-Address = 209.100.92.206 Framed-IP-Address = 209.100.92.207 Framed-IP-Address = 209.100.92.208 Framed-IP-Address = 209.100.92.209 Framed-IP-Address = 209.100.92.210 Framed-IP-Address = 209.100.92.211 Framed-IP-Address = 209.100.92.212 Framed-IP-Address = 209.100.92.213 Framed-IP-Address = 209.100.92.214 Framed-IP-Address = 209.100.92.215 Framed-IP-Address = 209.100.92.216 Framed-IP-Address = 209.100.92.217 Framed-IP-Address = 209.100.92.218 ========================================================================= 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

> What is "pmwho"? PMWHO is an application that connects to a Livingston Portmaster-like NAS and shows who is online by doing a 'show sessions', stripping out the page breaks and such, and outputting it to stdout. It has been known to work on USRobotics units with some modification. > ~~~~~~~~~ ~~~~~~~~~~~ > Charles Sprickman Internet Channel > INCH System Administration Team (212)243-5200 > spork@inch.com access@inch.com Timothy Brown [tcb@star-nets.com] StarNet, Inc. An Atlanta, GA Internet Service Provider

On Mon, 12 Jan 1998, Charles Sprickman wrote: > > Freeware C program written for Liv PM's that telnets into the box and gets > > connected user information.. Developed since neither the PM or the Netserver > > can supply that kind of information via SNMP.. > > > > Many have modified it to do thier own bidding as well.. There is also a perl > > program out there that does the same thing.. > > Is there any way to do this without shooting your "!root" all over the > place? > > Charles > I wrote an expect script so the only place to view the !root password is in the expect file. Set perms on the file heavy enough that if someone gets to this file, we've got more serious security probs. The thing I don't like about it is it's in clear text. I can't drop to interactive mode to type it in because automated utils also use this. -------------cut pmcmd.exp------------- #!/usr/local/bin/expect -- # # Telnet login to <portmaster> console # # Usage: pmcmd.exp terminal_server command # Eg. pmcmd.exp tc1.dyer.jorsm.com show sessions # # Copyright 1995 JORSM Enterprises, Inc. # Unlimited distribution allowed, but you can't hold # us responsible for anything. # # Please supply updates to jeff@jorsm.com # # Protect the perms on this file. Your !root password is stored # in clear text. # set timeout 10 match_max 80 set pm [lrange $argv 0 0] set cmd [lrange $argv 1 end ] # open a portmaster telnet session spawn telnet expect "telnet>" send "set escape ~\r" expect "telnet>" send "open $pm\r" expect "ogin:" send "!root\r" expect "word:" send "PASSWORD\r" expect ">" send "$cmd\r" expect { "Press Return for More " {send "\r"; exp_continue;} ">" {send "\r"; exit;} } -------------end pmcmd.exp------------- Now build your pmwho script -------------cut pmwho---------------- # you can | grep ESTAB to limit the output /usr/local/sbin/pm/pmcmd.exp $1 sho ses -------------end pmwho--------------- Hope this helps. ========================================================================= 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

> Freeware C program written for Liv PM's that telnets into the box and gets > connected user information.. Developed since neither the PM or the Netserver > can supply that kind of information via SNMP.. > > Many have modified it to do thier own bidding as well.. There is also a perl > program out there that does the same thing.. Is there any way to do this without shooting your "!root" all over the place? Charles > > -M > > `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' > 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. >

On Mon, 12 Jan 1998, Michael Wronski wrote: > At 02:28 PM 1/12/98 -0600, you wrote: > >I've had quite a few "complaints" that after installing our > >Enterprise Hubs, there is a noticably longer delay during > >the Windows 95 "Verifying username/password" sequence. I've > >check the radius accounting logs for possible symptoms, > >like increased Acct-Delay for trouble communicating with the > >radius server (auth and acct on same machine). > > > >Any suggestions? > > > Make sure your clients dont have "log into network" checked and that they > are only configured for the protocols that you support... Usually > unchecking IPX and NETBEUI for the typical ISP setup. This will reduce the > LCP negotiation time.. > We do...sorry, should have mentioned that. Also unchecked "log on to network" ========================================================================= 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

At 04:09 PM 1/12/98 -0500, you wrote: >> Freeware C program written for Liv PM's that telnets into the box and gets >> connected user information.. Developed since neither the PM or the Netserver >> can supply that kind of information via SNMP.. >> >> Many have modified it to do thier own bidding as well.. There is also a perl >> program out there that does the same thing.. > >Is there any way to do this without shooting your "!root" all over the >place? > No. -M

On Mon, 12 Jan 1998, Jordyn A. Buchanan wrote: > This is the correct behavior. The Netserver 16 does count the console > port, as you speculate, so it allocates a pool of 17 addresses. It would > be really nice if you could declare the size of the address pool, as you > can do with ComOS these days, especially since it's mentally easier to > think about blocks of 16 addresses as opposed to 17, at least for me. My > brain seems to work mostly in binary these days... > > Jordyn Interesting, I wasn't aware I could initiate a framed session over the console port. Thanks for verifying this. ========================================================================= 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

At 3:11 PM -0600 1/12/98, Jeff Lynch wrote: >On Mon, 12 Jan 1998, Michael Wronski wrote: > >Is there a bug in Netserver 16I dynamic IP address assignment that >causes it to assign address beyond the range of Start_address + Num_ports >For example: > >Netserver 16I assigned address: x.x.x.202 >Assigns addresses up to: x.x.x.218 > >Heres a grep | sort -u on IP pool addresses assigned. >This 16 port machine thinks it's IP address pool size is 17! >Maybe they're counting the console port? > > Framed-IP-Address = 209.100.92.202 > Framed-IP-Address = 209.100.92.203 > Framed-IP-Address = 209.100.92.204 > Framed-IP-Address = 209.100.92.205 > Framed-IP-Address = 209.100.92.206 > Framed-IP-Address = 209.100.92.207 > Framed-IP-Address = 209.100.92.208 > Framed-IP-Address = 209.100.92.209 > Framed-IP-Address = 209.100.92.210 > Framed-IP-Address = 209.100.92.211 > Framed-IP-Address = 209.100.92.212 > Framed-IP-Address = 209.100.92.213 > Framed-IP-Address = 209.100.92.214 > Framed-IP-Address = 209.100.92.215 > Framed-IP-Address = 209.100.92.216 > Framed-IP-Address = 209.100.92.217 > Framed-IP-Address = 209.100.92.218 This is the correct behavior. The Netserver 16 does count the console port, as you speculate, so it allocates a pool of 17 addresses. It would be really nice if you could declare the size of the address pool, as you can do with ComOS these days, especially since it's mentally easier to think about blocks of 16 addresses as opposed to 17, at least for me. My brain seems to work mostly in binary these days... Jordyn |----------------------------------------------------------------| |Jordyn A. Buchanan mailto:jordyn@bestweb.net | |Bestweb Corporation http://www.bestweb.net | |Senior System Administrator +1.914.271.4500 | |----------------------------------------------------------------|

> Make sure your clients dont have "log into network" checked "Log into network" should NOT be checked, unless the user needs to log onto a domain controller on the ISPs network. For most ISPs, this box should never be checked. -- Brad Wilson, student of Objectivism ------------ bradw at pobox.com --

Yes, I've noticed this as well, not counting the 'logging onto network' and ipx/netbuie options (all unchecked). ISDN seems to verify the username/password much faster than an x2 call.. it can't be a bandwidth issue, since the difference isnt that much. Where might the additional delay be coming from? - lv On Mon, 12 Jan 1998, Jeff Lynch wrote: > I've had quite a few "complaints" that after installing our > Enterprise Hubs, there is a noticably longer delay during > the Windows 95 "Verifying username/password" sequence. I've > check the radius accounting logs for possible symptoms, > like increased Acct-Delay for trouble communicating with the > radius server (auth and acct on same machine). > > Any suggestions? > > ========================================================================= > 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 > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

> > Brian, whoever you are, thanks much! I am guessing that you are referring to brian mcintire @ 3com. I spoke to him mostly (and he couldn't help me with my problem), but he tried hard; attempted to escalate the issue to upper levels; and even was genuinely hurt when we got his manager and the salesperson involved. Brian is definately one of the "good guys" [1] Thankfully, this list came through with the correct answers and 2 fully populated TCM racks and a RADIUS server did not become dumpster dust in the midwest. Never did get a call back from the 2nd level support people. doug douglas o'flaherty douglas@leftbank.com VP Ops http://www.leftbank.com/ The Left Bank Operation, Inc. TCP/IP Internetworking [1] -- no sacrasm intended.

> >Is there a bug in Netserver 16I dynamic IP address assignment that > >causes it to assign address beyond the range of Start_address + Num_ports > >For example: > > This is the correct behavior. The Netserver 16 does count the console > port, as you speculate, so it allocates a pool of 17 addresses. Does the Total Control hub do something similar? It doesn't seem to, since I've never had a problem with set limit 48 which I think (although yesterday, I knew it for sure!) allocates a pool of 48 addresses beginning with the "set assign" address. I never noticed this problem with the Netserve 16s. I have always assigned pools of 16 addresses to them, as: set ip addr x.y.z.1 set assign x.y.z.16 set ip addr x.y.z.2 set assign x.y.z.32 set ip addr x.y.z.3 set assign x.y.z.48 set ip addr x.y.z.4 set assign x.y.z.64 [...] Does this result in possible overlapping IP assignments between successive Net 16s? Do I have to worry about my TC hubs? --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."

> > 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). I found David Bolen's response quite enlightening, and I'll have to study it further. As he mentioned, there are simpler ways to do things if you're only concerned with active PPP logins. We use the ifNumber tree in the Netserver card, and only require two values. This works on our Net8s, Net16s and TCs - but we don't have ISDN or HDMs (yet). The algorithm is simply ## Take the number of ports, and subtract one for the ethernet users = ifNumber - 1 ## subtract 1 if console is active if (ifNumber.ifIndex.2 ne '(null)') users -= 1 That being said, I'm going to try write some code that implements David's suggestions since it seems more accurate. In particular, "active PPP logins" may be the same as "current users," but it's not quite the same as "modems in use." For instance, it doesn't consider modems negotiating new connections. --phil "All my life, I always wanted to be somebody. Now I see that I should have been more specific."

At 06:50 PM 1/12/98 -0500, you wrote: >Yes, I've noticed this as well, not counting the 'logging onto network' >and ipx/netbuie options (all unchecked). ISDN seems to verify the >username/password much faster than an x2 call.. it can't be a bandwidth >issue, since the difference isnt that much. Where might the additional >delay be coming from? > >- lv > >On Mon, 12 Jan 1998, Jeff Lynch wrote: > >> I've had quite a few "complaints" that after installing our >> Enterprise Hubs, there is a noticably longer delay during >> the Windows 95 "Verifying username/password" sequence. I've >> check the radius accounting logs for possible symptoms, >> like increased Acct-Delay for trouble communicating with the >> radius server (auth and acct on same machine). >> >> Any suggestions? >> Can any of you e-mail me a PPP trace of one of these "long" negotiations, it might show something... It is possible that there is something causing a lot of LCP negotiation.. It could be as simple as options in the RADIUS user config that cant be negotiated by some clients.. This could cause extra time spent in LCP.. -M

It's time for the user to call the local telco for a line quality check. As with all of the new protocols, the modems are constantly renegotiating the best line speed. I had a user call me up screaming in my ear about a similiar problem, it turns out he had a Bannana Slug doing the 90 volt slither in his telco junction box. (Yes Virginia they do conduct) Anyway, once the mess was cleaned up, his lines speed came up to our expected norms. The other suspect would again be on the user's end. Is it an external modem? And if so, is it connected to a 16550 compatible UART. Older generation UARTs have a max intrnal speed of 19200 bps, anything faster causes erratic behavior and data loss/retransmits. Again this is consistent with the symptoms you related. > Date: Sat, 10 Jan 1998 09:05:33 -0500 (EST) > From: System Administrator <sysadmin@evcom.net> > To: USRobotics TC Mailing List <usr-tc@xmission.com> > Subject: (usr-tc) ftp/tcp performance > Reply-to: usr-tc@lists.xmission.com > Over the past YEAR, I have had ONE customer (that I am aware of) > experience the problem below. Now I have another one. My "gut instinct" > tells me that this is not related to usr-tc, however I thought I would ask > here anyway because the clue factor seems to be quite high on this list. > :-) > > A customer begins an FTP transfer from from a local server on our network > to their home win95 system. In this case the customer has a solid x2 > connection. Initially they experience a throughput of approx. 5KB/s > (reasonable for x2), however the throughput continuously degrades over the > period of the transfer until they are getting an average of 1KB/s. I have > had the customer install the lastest MS DUN, with no positive results. > > I *cannot*, repeat, *cannot* duplicate the problem. Over my personal ISDN > I get a consistant 14KB/s. Over a test x2 connection I am able to sustain > throughput of 4-5KB/s using the SAME FTP client that the customer is > using, transfering the same file from the same FTP server. The customer's > modem is internal, and we have fiddled with every setting imaginable to > man (baud rates, ip header compression, etc, etc), with no success. As I > said before, the almost EXACT same thing happened with another customer in > 1997. > > Anyone ever seen this before, or have any suggestions? > > Regards, > > Jesse Sipprell > Senior Systems Engineer > Evolution Communications, Inc. > > * Finger sysadmin@evcom.net for my PGP Public Key * > Dave Sherry aka... Malevil@wclynx.com Webmaster@wclynx.com Postmaster@wclynx.com support@ap.net (707) 887-4640 (voice) (707) 887-1810 (fax) "Computers don't just die, They're murdered!"

Has anyone gotten the Solaris version of TCM to work through a Socks firewall? Using the runsocks wrapper doesn't seem to do the trick. -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1

At 04:30 PM 1/13/98 -0500, you wrote: > >What netserver command do I use to see what speeds people are >connecting to our terminal server at? > >I've tried "show all" and "show s12" for instance but those >don't seem to have the connect speed listed anywhere. > This information is only availalbe through TCM.. The netserver has no concept of connect speeds of the modems.. -M

What netserver command do I use to see what speeds people are connecting to our terminal server at? I've tried "show all" and "show s12" for instance but those don't seem to have the connect speed listed anywhere. I am wanting to see how they compare to our PM3's. Thanks, 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. =============================================================================

does anybody know of a way to set a feature enable key on multiple chassis. i've played around with xtcmfeat* in tcmunix which in the help menu, tries to explain a method of making a file? and then setting no target? has anybody used this option? thanks ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________|

On Tue, 13 Jan 1998, Eric J. Merkel wrote: > > What netserver command do I use to see what speeds people are > connecting to our terminal server at? > > I've tried "show all" and "show s12" for instance but those > don't seem to have the connect speed listed anywhere. > > I am wanting to see how they compare to our PM3's. > You don't. To find the connect speed you have to get the information from the NMC. Open TCM and click the RN/FL light on your first modem card so that all four channels are highlighted. Then go to View/Select All. Click the button that looks like a speedometer and you will see a window with the several traps listed. The ones you need are current receive and current transfer. Once you have those highlighted, click Add and then click OK. It will display the port speeds for all the ports you highlighted. You can do all the ports, or just highlight specific ports. :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::

On Tue, 13 Jan 1998, Brent Jay wrote: > You don't. To find the connect speed you have to get the information from > the NMC. Open TCM and click the RN/FL light on your first modem card so > that all four channels are highlighted. Then go to View/Select All. > Click the button that looks like a speedometer and you will see a window > with the several traps listed. The ones you need are current receive and > current transfer. Once you have those highlighted, click Add and then > click OK. It will display the port speeds for all the ports you > highlighted. You can do all the ports, or just highlight specific ports. > Thanks for the tip on quickly selecting all modems. What does the number on parenthesis mean after the trans/rec speed. eg. bps48000(42) ^^^^ ========================================================================= 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

On Tue, 13 Jan 1998, Jeff Lynch wrote: > On Tue, 13 Jan 1998, Brent Jay wrote: > > Thanks for the tip on quickly selecting all modems. What does the > number on parenthesis mean after the trans/rec speed. eg. > bps48000(42) > ^^^^ That's just the integer SNMP response from the NMC. You can find them all in the mdm.mib file. -CH

Does this work the same on HDM's ? Robert von Bismarck Petrel Communication S.A. -----Original Message----- From: Jeff Lynch [SMTP:jeff@mercury.jorsm.com] Sent: mercredi, 31. d=E9cembre 1997 20:13 To: usr-tc@lists.xmission.com Subject: Re: (usr-tc) Carrier Loss Delay Thanks. This looks even a little easier than setting Transparent under DS0 Status Types which I am currently doing upon suggestion from this thread. Happy New Year. =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D 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 On Wed, 31 Dec 1997, Clayton Zekelman wrote: >=20 > Open up a chassis with TCM > Select the first LED on the Dual T1 card. > Configure > Actions/Commands... > Select DS0 > Select the channels you want to busy out > Select OK > Change "No Command" to "Soft Busy Out" > Select Execute >=20 > Wait for the callers to log out. > Perform whatever maintenance you need to on the quads. > Repeat the above procedure, except this time select Restore instead of > Soft Busy Out. >=20 > A soft busy out will stop the channel from receiving calls, but will let > the current call stay up 'till the user is done. >=20 >=20 > You can also do this by plugging an RS-232 terminal into the back of the > Dual T1 Card. >=20 >=20 >=20 > At 10:22 AM 12/31/97 -0600, you wrote: > >On Wed, 31 Dec 1997, Clayton Zekelman wrote: > >>=20 > >> Could you take one span down at a time during non-peak periods? Thats > >> what we used to do when we only had one chassis. You can even soft busy > >> out individual DS0's if you want. =20 > > > >Hmmmm......what exactly is "soft busy". > > > >=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =3D=3D > >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 > > > > > >- > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. >=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.

Any word on when TCM for Solaris 2.6 will be out? I would think, that as with other applications, to recompile for Solaris 2.6 would be very trivial and shouldn't take long at all...yet its not out there yet! I know I oculd run Solaris 2.5.1, but I run 2.6 server on our admin machine. 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 -----------------------------/

TCM should run on 2.6 without any troubles, it doesn't require any kernel intervention perse so ... On Tue, Jan 13, 1998 at 08:06:16PM -0600, Brian made some electrons appear in the following form: > Any word on when TCM for Solaris 2.6 will be out? I would think, that as > with other applications, to recompile for Solaris 2.6 would be very > trivial and shouldn't take long at all...yet its not out there yet! I > know I oculd run Solaris 2.5.1, but I run 2.6 server on our admin machine. > > 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. --

On Tue, 13 Jan 1998, Brett Hawn wrote: > TCM should run on 2.6 without any troubles, it doesn't require any kernel > intervention perse so ... > TCM does NOT work on 2.6, which is why I sent my message out originally. Brian > On Tue, Jan 13, 1998 at 08:06:16PM -0600, Brian made some electrons appear in the following form: > > Any word on when TCM for Solaris 2.6 will be out? I would think, that as > > with other applications, to recompile for Solaris 2.6 would be very > > trivial and shouldn't take long at all...yet its not out there yet! I > > know I oculd run Solaris 2.5.1, but I run 2.6 server on our admin machine. > > > > 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. > > -- > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

Is there a way to make the radius or the NetServer refuse the request for ipx/netbuie instead of taking the time to time out? This would speed up the connection time for the users that don't know enough to uncheck this or to call in and ask our tech support about it. Who knows how many customers we may have lost due to a long conection time. Randy Doran CyberGate Network Operations On Tue, 13 Jan 1998, Michael Wronski wrote: > At 06:50 PM 1/12/98 -0500, you wrote: > >Yes, I've noticed this as well, not counting the 'logging onto network' > >and ipx/netbuie options (all unchecked). ISDN seems to verify the > >username/password much faster than an x2 call.. it can't be a bandwidth > >issue, since the difference isnt that much. Where might the additional > >delay be coming from? > > > >- lv > > > >On Mon, 12 Jan 1998, Jeff Lynch wrote: > > > >> I've had quite a few "complaints" that after installing our > >> Enterprise Hubs, there is a noticably longer delay during > >> the Windows 95 "Verifying username/password" sequence. I've > >> check the radius accounting logs for possible symptoms, > >> like increased Acct-Delay for trouble communicating with the > >> radius server (auth and acct on same machine). > >> > >> Any suggestions? > >> > > Can any of you e-mail me a PPP trace of one of these "long" negotiations, > it might show > something... > > It is possible that there is something causing a lot of LCP negotiation.. It > could be as simple as options in the RADIUS user config that cant be > negotiated by some > clients.. This could cause extra time spent in LCP.. > > -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. > >

At 09:51 PM 1/13/98 -0500, you wrote: >Is there a way to make the radius or the NetServer refuse the request for >ipx/netbuie instead of taking the time to time out? This would speed up >the connection time for the users that don't know enough to uncheck this >or to call in and ask our tech support about it. Who knows how many >customers we may have lost due to a long conection time. If the user is not configured for a protocol in RADIUS or if the protocol is unsupported the NAS will always NAK the config requests from the cleint.. The problem here is you dont know how many times the client will attemt the config request before giving up.. Some clients can be very persistant in their attempts to negotiate protocols in LCP. Win95 is the problem here.. There really is nothing the NAS can do to speed up the process. It must respond to all config requests that it gets from a client.. Changing this would break the RFC and could cause all kinds of other problems.. -M

On Tue, 13 Jan 1998, Randy Doran wrote: > Is there a way to make the radius or the NetServer refuse the request for > ipx/netbuie instead of taking the time to time out? This would speed up > the connection time for the users that don't know enough to uncheck this > or to call in and ask our tech support about it. Who knows how many > customers we may have lost due to a long conection time. > I don't think so since it is Windows 95 trying to establish those protocols and not the netserver. The netserver is not querying the client for Netbeui, so how can it do anything about it? Someone correct me if I'm wrong, but I don't see how the netserver could do anything about it. > > On Tue, 13 Jan 1998, Michael Wronski wrote: > > > At 06:50 PM 1/12/98 -0500, you wrote: > > >Yes, I've noticed this as well, not counting the 'logging onto network' > > >and ipx/netbuie options (all unchecked). ISDN seems to verify the > > >username/password much faster than an x2 call.. it can't be a bandwidth > > >issue, since the difference isnt that much. Where might the additional > > >delay be coming from? > > > > > >- lv > > > > > >On Mon, 12 Jan 1998, Jeff Lynch wrote: > > > > > >> I've had quite a few "complaints" that after installing our > > >> Enterprise Hubs, there is a noticably longer delay during > > >> the Windows 95 "Verifying username/password" sequence. I've > > >> check the radius accounting logs for possible symptoms, > > >> like increased Acct-Delay for trouble communicating with the > > >> radius server (auth and acct on same machine). > > >> > > >> Any suggestions? > > >> > > > > Can any of you e-mail me a PPP trace of one of these "long" negotiations, > > it might show > > something... > > > > It is possible that there is something causing a lot of LCP negotiation.. It > > could be as simple as options in the RADIUS user config that cant be > > negotiated by some > > clients.. This could cause extra time spent in LCP.. > > > > -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. > > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > :::::::::::::::::::::::::::::::::::: :: :: :: bjay@ionet.net :: :: ioNET network specialist :: :: break out the blender and :: :: mix me a spam margarita! :: :: 1-800-360-5183 405-270-0999 :: :: :: ::::::::::::::::::::::::::::::::::::

On Wed, 14 Jan 1998, Brent Jay wrote: > On Tue, 13 Jan 1998, Randy Doran wrote: > > > Is there a way to make the radius or the NetServer refuse the request for > > ipx/netbuie instead of taking the time to time out? This would speed up > > the connection time for the users that don't know enough to uncheck this > > or to call in and ask our tech support about it. Who knows how many > > customers we may have lost due to a long conection time. > > > > I don't think so since it is Windows 95 trying to establish those > protocols and not the netserver. The netserver is not querying the client > for Netbeui, so how can it do anything about it? Someone correct me if > I'm wrong, but I don't see how the netserver could do anything about it. In that case it was still the hope that the netserver could somehow send the client a DONT or WONT type of message to not handle those protocols.. - lv > > On Tue, 13 Jan 1998, Michael Wronski wrote: > > > > > At 06:50 PM 1/12/98 -0500, you wrote: > > > >Yes, I've noticed this as well, not counting the 'logging onto network' > > > >and ipx/netbuie options (all unchecked). ISDN seems to verify the > > > >username/password much faster than an x2 call.. it can't be a bandwidth > > > >issue, since the difference isnt that much. Where might the additional > > > >delay be coming from? > > > > > > > >- lv > > > > > > > >On Mon, 12 Jan 1998, Jeff Lynch wrote: > > > > > > > >> I've had quite a few "complaints" that after installing our > > > >> Enterprise Hubs, there is a noticably longer delay during > > > >> the Windows 95 "Verifying username/password" sequence. I've > > > >> check the radius accounting logs for possible symptoms, > > > >> like increased Acct-Delay for trouble communicating with the > > > >> radius server (auth and acct on same machine). > > > >> > > > >> Any suggestions? > > > >> > > > > > > Can any of you e-mail me a PPP trace of one of these "long" negotiations, > > > it might show > > > something... > > > > > > It is possible that there is something causing a lot of LCP negotiation.. It > > > could be as simple as options in the RADIUS user config that cant be > > > negotiated by some > > > clients.. This could cause extra time spent in LCP.. > > > > > > -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. > > > > > > > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the message. > > For information on digests or retrieving files and old messages send > > "help" to the same address. Do not use quotes in your message. > > > > > > :::::::::::::::::::::::::::::::::::: > :: :: > :: 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. >

We have a total control hub thats' not reporting the "connection failure" to the radius security/accounting server. I have set the trap settings to "enable logs" on the "on connection failure", then saved to the NVRam, but a few days later, it changes back to "disable all". This unit has been running for 35 days and hasn't been rebooted. Any clue/suggestion is much appreciated. Liping Chen -- Netsol Technologies 805 S. Lemon Ave. Walnut, CA 91789 (909) 869-6455 (909) 869-6459 fax Liping@netsol.net

This 30 page guide was mentioned on the list a while back, and I had a hell of a time finding it on 3Com/USR's site. To save time for others who may want it .... Go to http://totalservice.usr.com Login Select "Documentation Library" from the left column Search for x2chassisref.pdf It appears to be a pretty good overview, except that it hasn't been updated so that the chart on p. 24 lists the current versions of code. 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 _____________________________________________

Tom Holderby said once upon a time: > >Has anyone successfully used a Cisco 760 ISDN router to connect to your USR >TC Hub? If so I'd really appreciate it if you could email me a copy of the >configuration. I can get it to connect, and I can ping the router, but I >can't get it to route packets through to the computer on the downstream >side of the 760. Under current code, Cisco 76x can not connect to Quads running in ISDN-Modem mode, nor can they connect to HDM cards. They can only connect if your ISDN is being processed by the Munich card on the Netserver (ie: GW Slot != 0). Here is a configuration that we use with our customers. There are a few errors in it, but for the most part it is accurate: # This is for an unbonded connection set systemname <xmission account name> set ppp multilink off set multidestination off set Bridging Off set SWITCH <to the switch type your phone provider is using> set 1 SPID <your spid number> set 1 DIRECTORYNUMBER <your directory number> set PHONE1 = <your directory number> set 2 SPID <your second spid number> set 2 DIRECTORYNUMBER <your second directory number> set PHONE2 = <your second directory number> set ppp authentication incoming pap set ppp password client <your xmission password> cd cd internal set ip address 1.1.1.1 set ip netmask 255.255.255.0 cd cd LAN set IP Address <your router IP address> set SUBNET <your assigned subnet size> set IP Routing On cd set user xmission set 1 number <9900900 or 4370900> set 1 auto on demand 1 threshold 0 demand 2 threshold 48 demand 1 duration 1 demand 2 duration 1 demand 1 source lan demand 1 source both set encapsulation ppp set speed auto set profile disconnect keep set ppp authentication outgoing none set ppp multilink off set multidestination off set timeout 60 set ip routing on set ip framing none set ip rip update off set bridging off set ip route destination 0.0.0.0/0 gateway 0.0.0.0 propogate on cost 1 set ip pat off cd

Hi everyone on this list ! Is the following possible in the older USR TC chassis, with the following cards: 1 Dual PRI card 12 Quad Analog/Digital Modem Cards 1 Netserver card 1 NMC Card 2 Power Supply Units Corresponding NICs ISDN PRI #1: Handles X2, V.34 and ISDN calls Analog calls are routed to the first 6 Quad Analog/Digital Modem cards ISDN PRI #2: Handles only ISDN calls. All these calls are routed to the Netserver Quad Analog/Digital Modem Cards #7-12: 24 POTS lines are connected to these modems via the NICs Leading to the following on this one unit: 23 X2/ISDN capable lines 23 Additional ISDN capable lines 24 V.34 capable lines I would appreciate all comments regarding this configuration, or similar ones. Thanks Ian Roy

Has anyone successfully used a Cisco 760 ISDN router to connect to your USR TC Hub? If so I'd really appreciate it if you could email me a copy of the configuration. I can get it to connect, and I can ping the router, but I can't get it to route packets through to the computer on the downstream side of the 760. Thanks in advance. Tom Holderby Sundial Inetnet Services

I mentioned a while back that we were having some real trouble with our new HiPer DSP (HDM) cards; our hunt group includes several T1s at the top and 64 old analog lines at the bottom. With one of our HDMs installed in slot 14 of our newest chassis running the newest code, it drops the T1 carrier when the first call comes in that's to be routed to an analog line. It then won't take any more calls, until it's reset. Installed in slot 2, it drops carrier, but subsequently accepts calls, after disconnecting all of its current calls, of course. (We've tried three of them to be sure it's not a hardware problem.) Our customers are not amused. Many of the fine people from USR/3Com have worked on the problem, but appear to have hit a wall of some sort. The HDMs don't work for us, and nobody seems to really know why. (Looks like I'll be installing a new TC hub on Friday.)

On Thu, 15 Jan 1998, Garry Shtern wrote: > Hi, > > I have a problem of the following nature. A lot of my users and I, > myself, have noticed a following thing. Every so often the connection > freezes for about 5-10 seconds and then goes back to nromal. During that > time none of the information is transmitted nor received back. I have an > external Courier v.everything and I have noticed that my MR light goes off > during that time. > Now, I have checked everything I could think of on my TC hubs, but I can > not figure out what is going on. I have made all of them upgrade to the > latest DUN and winsock from microsoft's site, but the problem persists. > > I wonder if anyone else is experiencing a similar issue and if anyone > knows how to resolve it. Thanks a lot. Does this happen with some specfic application? or just with any application? Say if you just dial up connect and just do a ping do you see a stop and go traffic ? krish > > -Garry > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Netlink Support Personnel said once upon a time: >I have a Netserver 8/I-modem that is currently only being used for 56K X2 >dial-in. > >I have a customer who has about 5 computers on a TCP/IP network and wants >Internet access for all of them....dedicated ISDN. > >I don't want to route a whole class C subnet to them...waste of IP >addresses. > >What is the best way to route a group of IPs , x.x.x.2 to x.x.x.10 for >example? Follow subnetting rules, and assign them a /29. In your above case, it would look something like this in your RADIUS file: username Authentication-Type = Unix-PW Framed-IP-Address = X.X.X.1 Framed-IP-Netmask = 255.255.255.248, Service-Type = Framed, Idle-Timeout = 600 This would give them X.X.X.1 for their gateway, and X.X.X.2 - X.X.X.6 for their machines, and X.X.X.7 for a broadcast. Make sure you've got RIPv2 running between your TC and your router, or else it won't pass the route when the connection is established. If you don't understand subnetting, get a grip on it now. It will become more and more important once you start dividing up networks.

I have a Netserver 8/I-modem that is currently only being used for 56K X2 dial-in. I have a customer who has about 5 computers on a TCP/IP network and wants Internet access for all of them....dedicated ISDN. I don't want to route a whole class C subnet to them...waste of IP addresses. What is the best way to route a group of IPs , x.x.x.2 to x.x.x.10 for example? Curt

Our Span line has 40 phone numbers (from xxx-xxx-4400 to xxx-xxx-4439). All phone lines were work last week. For some reasons, we power down the Total Control Hub during the weeken. Now, the phone# xxx-xxx-4400 doesn't work (getting fast busy signal), xxx-xxx-4401 to xxx-xxx-4439 are working fine. Any idea why only xxx-xxx-4400 doesn't work? Any help will appreciated. Thanks much. yong chen.

Jordyn A. Buchanan said once upon a time: >The problem here is that the Netserver 8/I doesn't speak RIPv2, unless >you're running the yucky NetServer Plus code. I presume there is only one Netserver 8/I, and the customer doesn't dial in to any other device. Just make a static route from your router to the Netserver 8/I, and you don't have to have any routing protocol running. For small networks this is preferable, but of course, doesn't scale well. >I end up using the following really unfortunate mechanism to route subnets >to our customers: > >username Authentication-Type = Unix-PW > Framed-IP-Address = X.X.X.1 > Framed-IP-Netmask = 255.255.255.255, > Framed-Route = "X.X.X.2 X.X.X.1 1", > Framed-Route = "X.X.X.3 X.X.X.1 1", > Framed-Route = "X.X.X.4 X.X.X.1 1", That is mighty ugly..

At 1:32 PM -0700 1/15/98, Pete Ashdown wrote: >Netlink Support Personnel said once upon a time: > >>I have a Netserver 8/I-modem that is currently only being used for 56K X2 >>dial-in. >> >>I have a customer who has about 5 computers on a TCP/IP network and wants >>Internet access for all of them....dedicated ISDN. >> >>I don't want to route a whole class C subnet to them...waste of IP >>addresses. >> >>What is the best way to route a group of IPs , x.x.x.2 to x.x.x.10 for >>example? > >Follow subnetting rules, and assign them a /29. In your above case, it >would look something like this in your RADIUS file: > >username Authentication-Type = Unix-PW > Framed-IP-Address = X.X.X.1 > Framed-IP-Netmask = 255.255.255.248, > Service-Type = Framed, > Idle-Timeout = 600 > >This would give them X.X.X.1 for their gateway, and X.X.X.2 - X.X.X.6 for >their machines, and X.X.X.7 for a broadcast. > >Make sure you've got RIPv2 running between your TC and your router, or else >it won't pass the route when the connection is established. > >If you don't understand subnetting, get a grip on it now. It will become >more and more important once you start dividing up networks. The problem here is that the Netserver 8/I doesn't speak RIPv2, unless you're running the yucky NetServer Plus code. I end up using the following really unfortunate mechanism to route subnets to our customers: username Authentication-Type = Unix-PW Framed-IP-Address = X.X.X.1 Framed-IP-Netmask = 255.255.255.255, Framed-Route = "X.X.X.2 X.X.X.1 1", Framed-Route = "X.X.X.3 X.X.X.1 1", Framed-Route = "X.X.X.4 X.X.X.1 1", ...and so on... I consider this to be a horrible way of handling the situation, but given that the Netserver 8 and Netserver 16 seem to be several years behind the rest of the INternet in terms of routing, there doesn't seem to be a better way. Jordyn |----------------------------------------------------------------| |Jordyn A. Buchanan mailto:jordyn@bestweb.net | |Bestweb Corporation http://www.bestweb.net | |Senior System Administrator +1.914.271.4500 | |----------------------------------------------------------------|

On Thu, 15 Jan 1998 yong_chen@nyed.uscourts.gov wrote: > > Our Span line has 40 phone numbers (from xxx-xxx-4400 to > xxx-xxx-4439). All phone lines were work last week. For some > reasons, we power down the Total Control Hub during the weeken. Now, > the phone# xxx-xxx-4400 doesn't work (getting fast busy signal), > xxx-xxx-4401 to xxx-xxx-4439 are working fine. Any idea why only > xxx-xxx-4400 doesn't work? Any help will appreciated. Thanks much. The DS0 for the 4400 could be OOS or LOOS If your using fixed assignment on the modems, the modem mapped to that DS0 may not be active on the packet bus. The modem may be set to "nic" for its DTE source Brian > > > yong chen. > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Thu, 15 Jan 1998, Garry Shtern wrote: > Hi, > > I have a problem of the following nature. A lot of my users and I, > myself, have noticed a following thing. Every so often the connection > freezes for about 5-10 seconds and then goes back to nromal. During that > time none of the information is transmitted nor received back. I have an > external Courier v.everything and I have noticed that my MR light goes off > during that time. > > Now, I have checked everything I could think of on my TC hubs, but I can > not figure out what is going on. I have made all of them upgrade to the > latest DUN and winsock from microsoft's site, but the problem persists. > > I wonder if anyone else is experiencing a similar issue and if anyone > knows how to resolve it. Thanks a lot. Try to disable x2, it sounds like x2 is retraining. Brian > > -Garry > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Fri, 16 Jan 1998, Robert von Bismarck wrote: > Hi, > > I've just received several Netserver/16 Plus to put in some > small remote POP's which have half a dozen BRI's instead of PRI. The > configuration looks remarkably similar to the HiPerARC. So I stuffed it > with the setup of one of my HiPerARCs which works, changed the IP's and > off I go. > > I just want to make sure whether that's all I need or if I have > to configure things differently, like the modems, or else. To make > things short, USR/3Com support was uneffective. So, if anyone has a > configuration sample for these machines so I can match it to mine. > > Thank you in advance, > > Robert von Bismarck > Petrel Communication S.A. > The NETServer 8/16 running 4.x code is similar to HiPer ARC. You need to configure the interfaces, and setup the Bri spids. Configure radius and other information like snmp etc. Other than that there is nothing special to 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. >

Hi, I have a problem of the following nature. A lot of my users and I, myself, have noticed a following thing. Every so often the connection freezes for about 5-10 seconds and then goes back to nromal. During that time none of the information is transmitted nor received back. I have an external Courier v.everything and I have noticed that my MR light goes off during that time. Now, I have checked everything I could think of on my TC hubs, but I can not figure out what is going on. I have made all of them upgrade to the latest DUN and winsock from microsoft's site, but the problem persists. I wonder if anyone else is experiencing a similar issue and if anyone knows how to resolve it. Thanks a lot. -Garry

Can you mix 70 amp and 45 amp power supplies in an enterprise hub?

We have multiple TC racks, our oldest is an enterprise hub and the new ones are not. The old hub hunts through the modems in a seemingly random fashion. Actually I think it rings modem 1 then 2 then 3 and so on until it get to 48 and then starts again. This is regardless of which ones are in use. The new racks use the first available modem. Is there a setting somewhere for this? I have looked through the Manger software and can find no difference any where or in the netserver software. We use T1's, not PRI. Terry Kennedy OlyPen, Inc.

On Fri, 16 Jan 1998, Douglas O'Flaherty wrote: > > OH boy, here I go again. > > Stumped 3 com tech support again: > > I am authenticating users using a 3com/USR RADIUS server on Unix. All users > dialing in as single hosts. my FRAME info is all correct. > I have only 48 modems. The 1st 59 users dial-in absolutely fine. > The 60th user gets assigned an IP of 0.0.0.0 *not good* > > Obviously, the pool of dynamically assigned IP addresses is being depleted. > In other words, IP addresses are not being reclaimed when someone hangs up. This problem was present in some version of NETServer code I am not sure which one. The problem has since been fixed. If you are using hint assigned feature on the NETServer this could happen. So first make sure that hint assigned is not turned on. The other way you could get into this problem is that if you have a small pool and more dialup ports. That is say you have 48 ports - but you may have only 47 address. > > The only "hint" I can find in _any_ USR documentation is on > page 9-9 of the Netserver Software Reference Manual where is mentions > "Termination-Action = Manage-Resources" > I had assumed this would only apply to when I wanted RADIUS to manage my > IP range, not if I wanted the Netserver to do so. > Terminate action is resource management related. Here the NETserver and the Radius talk back to each other about the resource ( IP adress etc ) they have and will keep track of the same - if one dies then the other will release all the held up address and vice versa. What version of code are you using on the NETServer? Also how many ports do you have active and how big is the pool krish > Asking tech support (1) what do they think is happening on the Netserver > (2) what does Manage-Resources mean in this case has resulted in blank > stares. {Alright, it's on the phone, but I really could here the blank stare.} > > > from 'show global' > > Assigned Address: 207.xx.yy.zz > Assigned Pool Size: 60 > > Error in syslog: > > Jan 13 21:58:12 my-tc1 assigned IP count(60) >= assigned IP limit(59) > > Selections from Default User in RADIUS: > > USER_SERVICE_TYPE: 2 > FRAMED_PROTOCOL: 1 > FRAMED_ADDRESS: 255.255.255.254 > FRAMED_NETMASK: 255.255.255.255 > FRAMED_ROUTING: 3 > FRAMED_FILTER_ID: > FRAMED_MTU: > FRAMED_COMPRESSION: 1 > > TERMINATION_ACTION: > > POOL_NAME: > > > doug > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

On Fri, 16 Jan 1998, Douglas O'Flaherty wrote: > krish: > > thanks for listening :) > > > This problem was present in some version of NETServer code I am not sure > > which one. The problem has since been fixed. > > If you are using hint assigned feature on the NETServer this could > > happen. So first make sure that hint assigned is not turned on. The > > other way you could get into this problem is that if you have a small > > pool and more dialup ports. That is say you have 48 ports - but you may > > have only 47 address. > > Supposed to be among the latest & greatest version: > > Command> version > U.S. Robotics > Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.6.28 > Build date: Oct 2 1997 > Build time: 14:42:06 > > Network Interface Card: Ethernet & Frame Relay Combination (26) > ISDN Interface Card : MUNICH32 (4) > Packet Bus Circuit : Enhanced > > S1-S52 are showing active. But the 1st card is management. pool is 48. > If I do a show all when the error happens no ore than 40% are actually > ESTABLISHED. Since you have only 48 modems and assuming that you have the modem card from slot 2-slot 12 do the following set s1-s4 inactive reset s1-s4 these ports are being used by the PRI/T1 card so do not set them to active And now do a show global. You sill see assigned address and you global Screen should say something like this Hint assigned: OFF also there is a pool size typically it is set to 1 more than active ports in your case it should be 49. If this does not fix the problem email me with the NETserver info. I will telnet in and have a look krish > > doug > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

On Fri, 16 Jan 1998, Terry Kennedy wrote: > We have multiple TC racks, our oldest is an enterprise hub > and the new ones are not. The old hub hunts through the modems > in a seemingly random fashion. Actually I think it rings modem 1 > then 2 then 3 and so on until it get to 48 and then starts again. This > is regardless of which ones are in use. The new racks use the first > available modem. Is there a setting somewhere for this? I have looked > through the Manger software and can find no difference any where or in > the netserver software. We use T1's, not PRI. Copyright (c) 3Com Corporation, 1995-1997 Dual T1 PRI Application Card Revision 3.0.2 (Card Id: 27) Boot Code Linked Date : Mon Dec 04 17:41:48 1995 Operation Code Linked Date: Fri Sep 05 12:11:37 1997 Main Menu 1 Command 2 Status 3 Card Configuration 4 Inbound Call Routing Configuration 5 Span Line 1 Configuration 6 Span Line 2 Configuration 7 SW Fault Manager Event Logging 8 Logout Enter menu selection and press Return. Menu Selection (1-8): Select 3 Card Configuration Current Setting 1 Save current Configuration to NVRAM 2 Restore NVRAM Configuration 3 Restore Default Configuration 4 Timing Source Priority Assignment Span-1=1 Span-2=2 5 Chassis Slot Device Configuration 6 Modem Routing Method Fixed Assignment 7 Configure Local Console Password 8 Change DS0 state on Quad Modem NAC action Disabled 9 Companding Code Configuration U-Law (NOTE: Changing configuration parameters may effect calls in progress.) Enter menu selection and press Return or press Esc to exit. Menu Selection (1-9): #6 is what your looking for, it allows you to change how the modems are answered. > > > 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. > /-------------------------- 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 -----------------------------/

Our t1 supplier controls this. We can choose between 1-48 or most idle line gets the call. -----Original Message----- >We have multiple TC racks, our oldest is an enterprise hub >and the new ones are not. The old hub hunts through the modems >in a seemingly random fashion. Actually I think it rings modem 1 >then 2 then 3 and so on until it get to 48 and then starts again. This >is regardless of which ones are in use. The new racks use the first >available modem. Is there a setting somewhere for this? I have looked >through the Manger software and can find no difference any where or in >the netserver software. We use T1's, not PRI. > > >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.

Pardon my cruddy perl as it was tossed together in a hurry and really sho= uld be using SNMP.pm but *shrug* #!/usr/local/bin/perl -- $stuff =3D /usr/local/bin/snmpget -v 1 $ARGV[0] $ARGV[1] ".iso.org.dod.in= ternet.private.enterprises.429.1.2.2.5"; ($other, $temp1) =3D split(/=3D/,$stuff); chop($temp1); #$temp1 =3D~ s/\D//g; $temp1 *=3D 1.8; $temp1 +=3D 32; $uptime =3D /usr/local/bin/snmpget -v 1 $ARGV[0] $ARGV[1] system.sysUpTim= e.0; chop $uptime; ($one, $two, $three, $four, $five, $six, $seven) =3D split(/ /,$uptime); print "$temp1\n"; print "0\n"; print "$five $six $seven\n"; print "$ARGV[0]\n"; this is meant for output into MRTG so your mileage may vary, it should be self evident what the variables are. On Fri, Jan 16, 1998 at 03:24:23PM +0100, Robert von Bismarck made some e= lectrons appear in the following form: > Hi, >=20 > I need to check the temperature of my hubs. I can do it via the TCM, bu= t > I'd like to do it via SNMP and graph it into MRTG with a cron job. I > checked the MIB, but I got lost in the depths of it... > Is there an SNMP line that I can query to get this info=A0? I have NMC > software version 5.0.8 and hardware 6.0. >=20 > Any pointers would be appreciated, >=20 > Robert von Bismarck > Petrel Communication S.A. >=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

--IMA.Boundary.061469488 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part No, you cannot. The 45 and 70 amp power supplies are for different chassis. The 45 works in the older, non-fantray chassis, and the 70 amp only works in the newer Fantray chassis. There is also a 130 amp power supply that is only for the newer chassis. Buster cc: --IMA.Boundary.061469488 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Can you mix 70 amp and 45 amp power supplies in an enterprise hub? - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the 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.061469488 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 4BF3EE20; Fri, 16 Jan 98 05:05:06 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id EAA05331; Fri, 16 Jan 1998 04:46:43 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0xt9QD-00068e-00; Fri, 16 Jan 1998 03:54:45 -0700 Received: from mail.olypen.com [208.200.248.2] by lists.xmission.com with smtp (Exim 1.73 #4) id 0xt9QB-00068P-00; Fri, 16 Jan 1998 03:54:43 -0700 Received: from modems.olypen.com(src addr [208.200.248.9]) (538 bytes) by mail.olypen.com via sendmail with P\:esmtp/R:inet_hosts/T:smtp (sender: <terry@olypen.com>) id <m0xt9QB-0004hoC@mail.olypen.com> for <usr-tc@lists.xmission.com>; Fri, 16 Jan 1998 02:54:43 -0800 (PST) (Smail-3.2.0.96 1997-Jun-2 #19 built 1997-Jun-30) Message-Id: <m0xt9QB-0004hoC@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.061469488--

This is the variable I query in mrtg for the temp.. 'UsrTemp' => '1.3.6.1.4.1.429.1.2.2.5', -Luke > > Hi, > > I need to check the temperature of my hubs. I can do it via the TCM, but > I'd like to do it via SNMP and graph it into MRTG with a cron job. I > checked the MIB, but I got lost in the depths of it... > Is there an SNMP line that I can query to get this info�? I have NMC > software version 5.0.8 and hardware 6.0. > > Any pointers would be appreciated, > > Robert von Bismarck > Petrel Communication S.A. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Then what is the point of X2 if I disable it? At 09:05 PM 1/15/98 -0600, Brian wrote: >On Thu, 15 Jan 1998, Garry Shtern wrote: > >> Hi, >> >> I have a problem of the following nature. A lot of my users and I, >> myself, have noticed a following thing. Every so often the connection >> freezes for about 5-10 seconds and then goes back to nromal. During that >> time none of the information is transmitted nor received back. I have an >> external Courier v.everything and I have noticed that my MR light goes off >> during that time. >> >> Now, I have checked everything I could think of on my TC hubs, but I can >> not figure out what is going on. I have made all of them upgrade to the >> latest DUN and winsock from microsoft's site, but the problem persists. >> >> I wonder if anyone else is experiencing a similar issue and if anyone >> knows how to resolve it. Thanks a lot. > >Try to disable x2, it sounds like x2 is retraining. > >Brian > > >> >> -Garry >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the 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. > > Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

Then what is the point of X2 if I disable it? At 09:05 PM 1/15/98 -0600, Brian wrote: >On Thu, 15 Jan 1998, Garry Shtern wrote: > >> Hi, >> >> I have a problem of the following nature. A lot of my users and I, >> myself, have noticed a following thing. Every so often the connection >> freezes for about 5-10 seconds and then goes back to nromal. During that >> time none of the information is transmitted nor received back. I have an >> external Courier v.everything and I have noticed that my MR light goes off >> during that time. >> >> Now, I have checked everything I could think of on my TC hubs, but I can >> not figure out what is going on. I have made all of them upgrade to the >> latest DUN and winsock from microsoft's site, but the problem persists. >> >> I wonder if anyone else is experiencing a similar issue and if anyone >> knows how to resolve it. Thanks a lot. > >Try to disable x2, it sounds like x2 is retraining. > >Brian > > >> >> -Garry >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the 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. > > Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 09:27 AM 1/15/98 -0600, Tatai SV Krishnan wrote: >On Thu, 15 Jan 1998, Garry Shtern wrote: > >> Hi, >> >> I have a problem of the following nature. A lot of my users and I, >> myself, have noticed a following thing. Every so often the connection >> freezes for about 5-10 seconds and then goes back to nromal. During that >> time none of the information is transmitted nor received back. I have an >> external Courier v.everything and I have noticed that my MR light goes off >> during that time. > >> Now, I have checked everything I could think of on my TC hubs, but I can >> not figure out what is going on. I have made all of them upgrade to the >> latest DUN and winsock from microsoft's site, but the problem persists. >> >> I wonder if anyone else is experiencing a similar issue and if anyone >> knows how to resolve it. Thanks a lot. > >Does this happen with some specfic application? or just with any >application? Say if you just dial up connect and just do a ping do you >see a stop and go traffic ? There is no specific application. It does not seem to be application related. I never did a ping during this period. But I am assuming that if the MR(Modem Ready) light goes off that implies that there is no connectivity. Also, during this time the modem lights on the bar(in win95) do not blink indicating there is no traffic coming or leaving my PC. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

--=====================_884990998==_ Content-Type: text/plain; charset="us-ascii" In response to the many requests on this list for information regarding connecting Cisco 700 series TA to a TC chassis, I have done some tests and have a working configuration and software combination. I am able to get a multilink connection to the Quad-I modems and to the Munich card. No new modem code is necessary for successful connections with the Munich. In brief: Cisco 760: 4.1 w/Fast Step setup 3Com Netserver: 3.6.28 Quad-I: 5.7.x & 5.8.x (**BETA CODE READ BELOW!**) In Detail: Hardware: Cisco 760 Cisco Software used: 4.1(2): Software Version c760-i.b.US 4.1(2) - Nov 12 1997 19:06:52 Cisco 766 ISDN Stack Revision US 2.10 (5ESS/DMS/NI-1) Copyright (c) 1993-1997 by Cisco Systems, Inc. All rights reserved. Software is used subject to software license agreement contained with this product. By using this product you agree to accept the terms of the software license. Hardware Configuration: DRAM: 1.5MB Flash: 0.5MB POTS: Type 0 (Rev. Unknown) NT1: Installed ROM: 2.1(1) The Cisco config is attached. I used the Cisco Fast Step config tool, available on Cisco's web site, to generate the config file. No changes were made to the config file created. NOTE: I did start with a "set def" to clear any previous configs from the 760. 3Com Hardware & Software

Michael Wronski said once upon a time: >In response to the many requests on this list for information regarding >connecting >Cisco 700 series TA to a TC chassis, I have done some tests and have a >working configuration and software combination. Thank you for your help and dedication Mike. Can you contact Pinal Patel and let her know that USR owns a Cisco 700 series for testing? She seems to be at a complete loss at how to solve my Cisco 700 to HDM problem. >I am able to get a multilink connection to the Quad-I modems and to the >Munich card. >No new modem code is necessary for successful connections with the Munich. >Quad Modem Hardware: SS & DS Quad Modem cards >Quad-I modem Code used: 5.7.x & 5.8.x. (BETA CODE) >*THIS IS BETA CODE* do not ask me where to find it if you are not a BETA >customer. >The new Quad-I code should be available on the Total Service Web site at >the end >of the January. You can probably see my question coming... What about HDM code? How closely related is the Quad-I code to the HDM code?

I have some further information on this. There is an engineering release for the Quads 5.7.91/5.8.91 that has this fixed in it, that should be available by mid next week. We would prefer that you wait for the general release, but if you need it urgently you can request the engineering release. mwronski@coredump.ae.usr.com on 01/16/98 10:49:58 AM Please respond to usr-tc@lists.xmission.com cc: In response to the many requests on this list for information regarding connecting Cisco 700 series TA to a TC chassis, I have done some tests and have a working configuration and software combination. I am able to get a multilink connection to the Quad-I modems and to the Munich card. No new modem code is necessary for successful connections with the Munich. In brief: Cisco 760: 4.1 w/Fast Step setup 3Com Netserver: 3.6.28 Quad-I: 5.7.x & 5.8.x (**BETA CODE READ BELOW!**) In Detail: Hardware: Cisco 760 Cisco Software used: 4.1(2): Software Version c760-i.b.US 4.1(2) - Nov 12 1997 19:06:52 Cisco 766 ISDN Stack Revision US 2.10 (5ESS/DMS/NI-1) Copyright (c) 1993-1997 by Cisco Systems, Inc. All rights reserved. Software is used subject to software license agreement contained with this product. By using this product you agree to accept the terms of the software license. Hardware Configuration: DRAM: 1.5MB Flash: 0.5MB POTS: Type 0 (Rev. Unknown) NT1: Installed ROM: 2.1(1) The Cisco config is attached. I used the Cisco Fast Step config tool, available on Cisco's web site, to generate the config file. No changes were made to the config file created. NOTE: I did start with a "set def" to clear any previous configs from the 760. 3Com Hardware & Software

Hi, I've just received several Netserver/16 Plus to put in some small remote POP's which have half a dozen BRI's instead of PRI. The configuration looks remarkably similar to the HiPerARC. So I stuffed it with the setup of one of my HiPerARCs which works, changed the IP's and off I go. I just want to make sure whether that's all I need or if I have to configure things differently, like the modems, or else. To make things short, USR/3Com support was uneffective. So, if anyone has a configuration sample for these machines so I can match it to mine. Thank you in advance, Robert von Bismarck Petrel Communication S.A.

Thus spake Chandru Bolaki >I have some further information on this. There is an engineering release >for the Quads 5.7.91/5.8.91 that has this fixed in it, that should be >available by mid next week. We would prefer that you wait for the general >release, but if you need it urgently you can request the engineering >release. Has what specifically fixed in it? If its a fix regarding being unable to make MLPPP calls, I definitely need to get my hands on it...I've had a trouble ticket open since Dec. 31 concerning this (27457). OBUSRTechSupportRant: this ticket has no less than 4 places in the notes where I am either requesting a callback or some other sort of communication from USR tech support...to this point, I have had to initiate *all* communications. To be fair to Vito, he did try to call, but had written down the wrong phone number (though the number was correct in the phone number field of the ticket as it shows up on totalservice) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

Douglas O'Flaherty said once upon a time: >FRAMED_ADDRESS: 255.255.255.254 >FRAMED_NETMASK: 255.255.255.255 Get rid of this. It isn't necessary for pooled addresses and is just plain wrong. I'll bet the pool will act normally once it is done.

Hi, I need to check the temperature of my hubs. I can do it via the TCM, = but I'd like to do it via SNMP and graph it into MRTG with a cron job. I checked the MIB, but I got lost in the depths of it... Is there an SNMP line that I can query to get this info=A0? I have NMC software version 5.0.8 and hardware 6.0. Any pointers would be appreciated, Robert von Bismarck Petrel Communication S.A.

On Friday 16 January 1998, at 15 h 24, the keyboard of Robert von=20 Bismarck <rvb@petrel.ch> wrote: > I need to check the temperature of my hubs. I can do it via the TCM, bu= t > I'd like to do it via SNMP and graph it into MRTG with a cron job. I > checked the MIB, but I got lost in the depths of it... grep is your friend.=20 grep -i -C temperature /usr/local/lib/MRTG/usr-mibs/* > Is there an SNMP line that I can query to get this info=A0? I have NMC > software version 5.0.8 and hardware 6.0. With MRTG (nice graph when the fan breaks): Target[tc-temp]: 1.3.6.1.4.1.429.1.2.2.5&1.3.6.1.4.1.429.1.2.2.5:thepassw= o rd@phone-admin=20 MaxBytes[tc-temp]: 100 Title[tc-temp]: phone-gw temperature Options[tc-temp]: gauge,nopercent YLegend[tc-temp]: Number of deg. C ShortLegend[tc-temp]: =B0 LegendO[tc-temp]: Legend1[tc-temp]: Number of =B0 C LegendI[tc-temp]: &nbsp;temp. Colours[tc-temp]: ROUGE#ff1111,ROUGE#ff1111,DARK GREEN#006600,VIOLET#ff00= f f PageTop[tc-temp]: <H1>Number of =B0 C on phone-gw </H1> <TABLE> <TR><TD>System:</TD><TD>phone-gw.internatif.org</TD></TR> <TR><TD>Maintainer:</TD><td>netadm@internatif.org</TD></TR> <TR><TD>Max Number:</TD> <TD>60</TD></TR> </TABLE>

From what I've gathered following the list and and experience with the RAS that the netserver has based it's PPP environment on, the IP address pool needs to be (1) greater than the total number of IP addresses that will be used by the modems. The requirement is due to one address being captured by the virtual machine of the RAS service, and it will always be the first address in the pool. Note - this is different from the network port address assigned to the network port interface. The error you listed below is an indication of what I described above. Out of 60 address only 59 are available to the modems due to the capture of the first address in the range by the virtual server. >> Error in syslog: >> Jan 13 21:58:12 my-tc1 assigned IP count(60) >= assigned IP limit(59) Also from reading your post, try setting the pool to only 49 addresses. By matching this and then running your burn in test, any problems should become clear quickly. Taking away the options forces the system to behave in accordance with the RFCs and the limits of the software. Any deviation should show in the error traps as they occur. The IP addresses should release imediately upon disconnect of the modem, and be ready for the next connection. There should not be any significant lag in availability between sessions. The only things that could cause such a lag is a socket being held open on the RAS server using that IP or the RAS server software itself is corrupt. In the case of corruption, a simple reload should fix the problem. > From: "Douglas O'Flaherty" <douglas@leftbank.com> > [snip] > I've tried with 48 (actual number of modems) and now with 60. It seems > more likely that reclaiming doesn't happen. Running tests *right now* > and We are finding some IP addresses are being recycled. > > now I am really unhappy. > > > > > For Douglas, since the error is occurring right on the boundary of > > what would normally be the pool maximum, is there any way that the > > NETServer might think that one of the ports you don't expect it to be > > counting towards the pool is actually a PPP session that has an > > address allocated? Perhaps the console port or something? > > > > Alternatively, is there anything stopping you from just defining the > > pool to be 61 addresses rather than 60 in case there's some other > > off-by-one error occurring? (noting that the log from the NETServer > > seems to imply it thinks the pool size is 59 rather than 60) > > > > -- David > > > > /-----------------------------------------------------------------------\ > > \ David Bolen \ Internet: db3l@ans.net / > > | ANS Communications, Inc. \ Phone: (914) 701-5327 | > > / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ > > \-----------------------------------------------------------------------/ > > Dave Sherry Site Network Administrator 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

We are finally ready to start the 8/16 beta for Pilgrim code. If you are registered you have received an e-mail message from me today.. If you didn't get a message from me regarding this beta and are interested in participating,please send me an e-mail(not to this list) and go to the total service web site and fill out a beta app.. I sent mail to everyone that I have an application from.. -Mike `'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`' Mike Wronski(mwronski@coredump.ae.usr.com) 3Com/U.S.Robotics Network Systems Engineer PGP: http://coredump.ae.usr.com/pgp (Prefered)

OH boy, here I go again. Stumped 3 com tech support again: I am authenticating users using a 3com/USR RADIUS server on Unix. All users dialing in as single hosts. my FRAME info is all correct. I have only 48 modems. The 1st 59 users dial-in absolutely fine. The 60th user gets assigned an IP of 0.0.0.0 *not good* Obviously, the pool of dynamically assigned IP addresses is being depleted. In other words, IP addresses are not being reclaimed when someone hangs up. The only "hint" I can find in _any_ USR documentation is on page 9-9 of the Netserver Software Reference Manual where is mentions "Termination-Action = Manage-Resources" I had assumed this would only apply to when I wanted RADIUS to manage my IP range, not if I wanted the Netserver to do so. Asking tech support (1) what do they think is happening on the Netserver (2) what does Manage-Resources mean in this case has resulted in blank stares. {Alright, it's on the phone, but I really could here the blank stare.} from 'show global' Assigned Address: 207.xx.yy.zz Assigned Pool Size: 60 Error in syslog: Jan 13 21:58:12 my-tc1 assigned IP count(60) >= assigned IP limit(59) Selections from Default User in RADIUS: USER_SERVICE_TYPE: 2 FRAMED_PROTOCOL: 1 FRAMED_ADDRESS: 255.255.255.254 FRAMED_NETMASK: 255.255.255.255 FRAMED_ROUTING: 3 FRAMED_FILTER_ID: FRAMED_MTU: FRAMED_COMPRESSION: 1 TERMINATION_ACTION: POOL_NAME: doug

Jeff What we know specifically has been fixed is an issue with the Cisco 760 unable to make a MLPPP connection to the Quads which was referred to in Mike Wronski's note. I looked up your ticket and it basically refers to about 10% of your calls unable to make MLPPP connections. There is no info logged on which devices are unable to make MLPPP connections. The ticket has been escalated to Tier 2. At this point, you should try terminating the calls on the Netserver munich like Mike suggests, and see what results you get. The Quad E.R. can be tried if you do not want to do that. Chandru jeffm@iglou.com on 01/16/98 12:51:51 PM Please respond to usr-tc@lists.xmission.com cc: Thus spake Chandru Bolaki >I have some further information on this. There is an engineering release >for the Quads 5.7.91/5.8.91 that has this fixed in it, that should be >available by mid next week. We would prefer that you wait for the general >release, but if you need it urgently you can request the engineering >release. Has what specifically fixed in it? If its a fix regarding being unable to make MLPPP calls, I definitely need to get my hands on it...I've had a trouble ticket open since Dec. 31 concerning this (27457). OBUSRTechSupportRant: this ticket has no less than 4 places in the notes where I am either requesting a callback or some other sort of communication from USR tech support...to this point, I have had to initiate *all* communications. To be fair to Vito, he did try to call, but had written down the wrong phone number (though the number was correct in the phone number field of the ticket as it shows up on totalservice) -- 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.

just FYI for all: Until this was set in the default RADIUS user, no IP addresses were being assigned. This list actually identified what the proper values for these fields. I am confident this information is correct. > > Douglas O'Flaherty said once upon a time: > > >FRAMED_ADDRESS: 255.255.255.254 > >FRAMED_NETMASK: 255.255.255.255 > > Get rid of this. It isn't necessary for pooled addresses and is just plain > wrong. > > I'll bet the pool will act normally once it is done. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Pete Ashdown <pashdown@xmission.com> writes: > Douglas O'Flaherty said once upon a time: > > >FRAMED_ADDRESS: 255.255.255.254 > >FRAMED_NETMASK: 255.255.255.255 > > Get rid of this. It isn't necessary for pooled addresses and is just plain > wrong. Actually, while the netmask might be superfluous, the address is technically required and correct, unless you want to assume default behavior of the NETServer. And for that matter, since several releases of NETServer code really want a dialup user netmask to be 255.255.255.255 to work properly, it's probably safer to include that in the response as well. The RADIUS attribute Framed-IP-Address (8) needs to be sent to the NETServer with a value of 255.255.255.254 to tell the NETServer (or any NAS using RADIUS for that matter) to pick an address from a locally defined pool. The other special value is 255.255.255.255 to be used to permit the dialup user to negotiate an address. I do believe that the NETServer may default to using an assigned pool if no address information is in the packet, but that might be a little risky since the other default it might have would be to negotiate with the user. I'm assuming that the dictionary in use is using the older attribute name (no "IP" in the middle) but that the above references attribute with the numeric value 8. > I'll bet the pool will act normally once it is done. For Douglas, since the error is occurring right on the boundary of what would normally be the pool maximum, is there any way that the NETServer might think that one of the ports you don't expect it to be counting towards the pool is actually a PPP session that has an address allocated? Perhaps the console port or something? Alternatively, is there anything stopping you from just defining the pool to be 61 addresses rather than 60 in case there's some other off-by-one error occurring? (noting that the log from the NETServer seems to imply it thinks the pool size is 59 rather than 60) -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | ANS Communications, Inc. \ Phone: (914) 701-5327 | / 100 Manhattanville Rd, Purchase, NY 10577 \ Fax: (914) 701-5310 \ \-----------------------------------------------------------------------/

[snip] I've tried with 48 (actual number of modems) and now with 60. It seems more likely that reclaiming doesn't happen. Running tests *right now* and We are finding some IP addresses are being recycled. now I am really unhappy. > > For Douglas, since the error is occurring right on the boundary of > what would normally be the pool maximum, is there any way that the > NETServer might think that one of the ports you don't expect it to be > counting towards the pool is actually a PPP session that has an > address allocated? Perhaps the console port or something? > > Alternatively, is there anything stopping you from just defining the > pool to be 61 addresses rather than 60 in case there's some other > off-by-one error occurring? (noting that the log from the NETServer > seems to imply it thinks the pool size is 59 rather than 60) > > -- 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. >

Once upon a time Ming Lu shaped the electrons to say... >Speak of experience, livingstone PM3 is cheap, but is is also cheap made. >I used them twice (my customers bought them). PM3 only support RIP and >limited OSPF functions, not even VLSM. Tech Support was very lousy too; >one of senior engineer told me that VLSM was not a part of OSPF >spec...:-). This is COMPLETE BULLSHIT. The PM-3 did OSPF *DAY ONE* with FULL VLSM and CIDR. Even OSPF NSSA, which many OSPF implementations lack. It has also supported BGP-4 for some time now, since ComOS 3.7. And I doubt very highly ANY senior engineer here ever told you that VLSM wasn't part of OSPF. >Take look at www.lanquest.com, search test reports reagrding >communication access servers. You will understand that why ascend market You mean the tests where the PM-3 beats the USR TC and Ascend MAX? <URL:http://www.livingston.com/Marketing/Products/lanquest_benchmark.shtml> >USR'S TOTALCONTROL is quite good product. It supports all of major >protocols, price vs. performance was good, but it does not have that >many functionalities as AS5200 has. This from a person who said the PM-3 doesn't do OSPF - when the USR doesn't. -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

Once upon a time Ming Lu shaped the electrons to say... >Speak of experience, livingstone PM3 is cheap, but is is also cheap made. >I used them twice (my customers bought them). PM3 only support RIP and >limited OSPF functions, not even VLSM. Tech Support was very lousy too; >one of senior engineer told me that VLSM was not a part of OSPF >spec...:-). This is COMPLETE BULLSHIT. The PM-3 did OSPF *DAY ONE* with FULL VLSM and CIDR. Even OSPF NSSA, which many OSPF implementations lack. It has also supported BGP-4 for some time now, since ComOS 3.7. And I doubt very highly ANY senior engineer here ever told you that VLSM wasn't part of OSPF. >Take look at www.lanquest.com, search test reports reagrding >communication access servers. You will understand that why ascend market You mean the tests where the PM-3 beats the USR TC and Ascend MAX? <URL:http://www.livingston.com/Marketing/Products/lanquest_benchmark.shtml> >USR'S TOTALCONTROL is quite good product. It supports all of major >protocols, price vs. performance was good, but it does not have that >many functionalities as AS5200 has. This from a person who said the PM-3 doesn't do OSPF - when the USR doesn't. -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

krish: thanks for listening :) > This problem was present in some version of NETServer code I am not sure > which one. The problem has since been fixed. > If you are using hint assigned feature on the NETServer this could > happen. So first make sure that hint assigned is not turned on. The > other way you could get into this problem is that if you have a small > pool and more dialup ports. That is say you have 48 ports - but you may > have only 47 address. Supposed to be among the latest & greatest version: Command> version U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.6.28 Build date: Oct 2 1997 Build time: 14:42:06 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Enhanced S1-S52 are showing active. But the 1st card is management. pool is 48. If I do a show all when the error happens no ore than 40% are actually ESTABLISHED. doug

Does anyone have reference to technical papers explaining how Net Medic is flawed? I recall great laughter was had at an ISP industry meeting when Net Medic was mentioned as a monitoring tool. Anyone recall when/where /what that meeting was? I am creating a form letter for our subscribers explaining why we refuse to support the "event report" Net Medic frequently mails us about our dire 'egress router' failures. I want to give 3COM's marketing crew a big hug for the headache they have caused us by promoting this poor product (or what we call around here "the fiction generator"). Below are the URLs I have looked across so far: http://www.vvm.com/whynonet.htm http://www.vvm.com/netmedic.htm http://www.isp-marketing.com/9707/threads.html#00169 (discussion topic) http://www.pcworld.com/news/daily/data/0497/970415140634.html http://www.vitalsigns.com/products/nm/ispletter.html http://www.rocketdownload.com/disc1/_homthr/0000000e.htm http://www.newstimes.com/archive/may0897/cpe.htm http://www.cs.bu.edu/groups/oceans/news/vs.html http://www.vitalsigns.com/support/nm/faqs.html#AutoCure_FAQs http://www.vitalsigns.com/support/nm/faqs.html#Data_Accuracy_FAQs http://www.vitalsigns.com/support/nm/faqs.html#Email_fp_FAQs http://www.vitalsigns.com/support/nm/faqs.html#NM_email_FAQs http://www.vitalsigns.com/support/nm/relnotes.html http://www.vitalsigns.com/support/nm/interpret/nmresults/quickref.html http://www.vitalsigns.com/support/modemtips.html http://www.vitalsigns.com/support/nmp/relnotespro12.html http://www.vitalsigns.com/products/index.html http://www.vitalsigns.com/products/vista/wp/characterizing.html Andy Dalrymple XMission Technical Support

On Sat, 27 Dec 1997, Jake Messinger wrote: ->On Fri, 26 Dec 1997, Erkan Alemdar wrote: -> ->> Max 4060, PM3, BAY 8000, AS3000 or USR ->> So many choices... -> ->Not really that many, yet... -> ->> We are an ISP in Turkey, currently we are upgrading our system from analog ->> to digital PIR lines. we will be installing around 1000-2000 ports in 98, ->> So we have to buy new access concentrators, but we are having difficulty ->> deciding which one to choose. They all say they are the best. ->> Price wise Ascend is the most expensive and Bay is the cheapest. What do you ->> recommend??? ->> What is the latest market share of those companies? -> ->I recommend PM3. They were bought by Lucent which makes them the largest ->Access Server company in the world. They are very good products, and have ->very good tech support and customer service, oeverseas as well as ->domestic. They seem to respond more openly and quickly to customer ->feedback/complaints/demands, etc... Speak of experience, livingstone PM3 is cheap, but is is also cheap made. I used them twice (my customers bought them). PM3 only support RIP and limited OSPF functions, not even VLSM. Tech Support was very lousy too; one of senior engineer told me that VLSM was not a part of OSPF spec...:-). I was in oversaes and had a couple of questions, livingston policy was callback within 24 hours. They never did, I called back and they claimed that the phone was busy...:-). then the guy could not answer my questions ans said that they would call back, guess what? they never did; two weeks later, I left that country. ->You should get info on the new PM 4. -> ->Speaking of market, Ascend stock has declined steadily over the past year ->from a high of 88 to as low as 20. Livingston had been privately held ->until the 650 mil buyout, but their sales pretty much doubled every year. ->Bay, of course, is made up of smaller companies that were bought, e.g. ->Synoptics, Performance Technologies, ... Take look at www.lanquest.com, search test reports reagrding communication access servers. You will understand that why ascend market is shrinking. Bay annex was quite good hardware, but software was very buggy (I used them in later 1996); I heard that tac support is not that good either, not much bad attitute, but qurlity of technical skills (I am sure that they have some good engineers) of TAC support. ->USR products do not support major routing protocols and are quite large ->and require expensive cooling fan arrays. USR'S TOTALCONTROL is quite good product. It supports all of major protocols, price vs. performance was good, but it does not have that many functionalities as AS5200 has. _ming ============================================================================ Ming Lu Email: mlu@hq.si.net Network Tech Consulting Engineer Phone: 703-689-5290 (w) Engineering Division 703-855-4194 (m) Global One Telecommunications, LLT. 703-689-6575 (f) ============================================================================ "Do not pay attention to every word people say, or you may hear your servant cursing you ---- for you know in your heart that many times you yourself have cursed others."

Another data point on the "why are my IP pools not being reclaimed under heavy load"? Assigned Address: xx.yy.zz.11 Assigned Pool Size: 49 No IPs below xx.yy.zz.34 are being handed out to the dial-in clients. all IPs above .34 to .59 have been handed out repeatedly. There must be some misconfiguration somewhere. doug

On Fri, 16 Jan 1998, MegaZone wrote: > Once upon a time Ming Lu shaped the electrons to say... > >Speak of experience, livingstone PM3 is cheap, but is is also cheap made. > >I used them twice (my customers bought them). PM3 only support RIP and > >limited OSPF functions, not even VLSM. Tech Support was very lousy too; > >one of senior engineer told me that VLSM was not a part of OSPF > >spec...:-). > > This is COMPLETE BULLSHIT. Its his opinion wrong as it may be. > > The PM-3 did OSPF *DAY ONE* with FULL VLSM and CIDR. Even OSPF NSSA, Oh that part. Are you sure it did ospf since day 1? There was some MAJOR feature that the very first pm 3's were missing but it was so lng ago I forgot. It must have been quickly added. > which many OSPF implementations lack. It has also supported BGP-4 for > some time now, since ComOS 3.7. The thing that got me, was that he said it was cheaply made. Ive been in this business for 20 years, and I can say that the PM 3 is VERY well made, much better even that the PM 2, especially if you compare it to the pipe 50 or the max. Im talking about physical construction here. You want to see a cheaply made unit, try a pipe 50. Weak plastic cases, no power switches, no port labels on the back, a flimsy switch that did nothing, etc... ~~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~ Jake Messinger ph:713-772-6690 Livingston Var AMS, Inc. fx:713-774-3498 Medical Billing 8300 Bissonnet #400 jake@ams.com Internet Services Houston, Texas 77074 www.ams.com/~jake Business Management ~~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~

Once upon a time Jake Messinger shaped the electrons to say... >> The PM-3 did OSPF *DAY ONE* with FULL VLSM and CIDR. Even OSPF NSSA, >Oh that part. Are you sure it did ospf since day 1? There was some MAJOR Completely. OSPF was in ComOS 3.5 - which shipped FIRST on the PM-3 at FCS. It followed on everything else shortly thereafter. ComOS 3.5 is the first OS for the PM-3 so the PM-3 has *never* been without OSPF. >feature that the very first pm 3's were missing but it was so lng ago I >forgot. It must have been quickly added. Perhaps you are thinking of CT1 dialin or BGP4 - which came officially in 3.7 (but were around in 3.5.1bx for months before that). I'm a bit sorry for the harsh tone in my first letter, but frankly I tend to get a bit bent when someone states completely false info as if it were a fact. I always strive to be as correct as possible and regret it when I do make an error. -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

Thus spake Chandru Bolaki >I looked up your ticket and it basically refers to about 10% of your calls >unable to make MLPPP connections. There is no info logged on which devices >are unable to make MLPPP connections. The ticket has been escalated to >Tier 2. Actually, the ticket is wrong, about 10% of our calls are *able* to make MLPPP connections, the other 90% are *un*able to make the MLPPP connections. We are seeing this problem with at least one of each of the following platforms, Ascend Pipeline 50, Bay Networks Instant Internet, and Vito (in your first level tech support) was able to reproduce the problem with a sportster 128k, and got an MLPPP log from NT. >At this point, you should try terminating the calls on the Netserver munich >like Mike suggests, and see what results you get. The Quad E.R. can be >tried if you do not want to do that. We have never switched to using the IModem code, have been using the Munich card all along. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

I am sorry that my statemenets regarding PM3 was not entirely correct. I was doing job with missing manuals and TAC support from livingston gave these "false" info. But my experience with TAC support was absolutely true, I had to finish my work with guessing (based on my knowledge of Cisco products, honorstly, PM3 user interface was quite similar with cisco and much better than bay's manaul driven user interface). _ming On Fri, 16 Jan 1998, MegaZone wrote: ->Once upon a time Jake Messinger shaped the electrons to say... ->>> The PM-3 did OSPF *DAY ONE* with FULL VLSM and CIDR. Even OSPF NSSA, ->>Oh that part. Are you sure it did ospf since day 1? There was some MAJOR -> ->Completely. OSPF was in ComOS 3.5 - which shipped FIRST on the PM-3 at ->FCS. It followed on everything else shortly thereafter. ComOS 3.5 is ->the first OS for the PM-3 so the PM-3 has *never* been without OSPF. -> ->>feature that the very first pm 3's were missing but it was so lng ago I ->>forgot. It must have been quickly added. -> ->Perhaps you are thinking of CT1 dialin or BGP4 - which came officially ->in 3.7 (but were around in 3.5.1bx for months before that). -> ->I'm a bit sorry for the harsh tone in my first letter, but frankly I tend ->to get a bit bent when someone states completely false info as if it were ->a fact. I always strive to be as correct as possible and regret it when ->I do make an error. -> ->-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 -> ============================================================================ Ming Lu Email: mlu@hq.si.net Network Tech Consulting Engineer Phone: 703-689-5290 (w) Engineering Division 703-855-4194 (m) Global One Telecommunications, LLT. 703-689-6575 (f) ============================================================================ "Do not pay attention to every word people say, or you may hear your servant cursing you ---- for you know in your heart that many times you yourself have cursed others."

Our dialin lines seem to be working fine. Plenty of 33.6/31.2 connects all around and good throughput. However, dialouts are severely limited. When a user telnets to a device for dialout, connects are limited to 19.2k. This seems to be directly related to the modem's reported DTE rate (19.2). For instance, an ATI5 returns: USRobotics Digital Quad NVRAM Settings... Copyright, 1988-97, U.S. Robotics. All rights reserved. DIAL=TONE B0 F1 X7 BAUD=19200 PARITY=N WORDLEN=8 &A3 &B1 &G0 &H1 &I0 &K1 &L0 &M4 &N0 &P0 &R2 &S0 &T4 &U0 &X0 &Y1 %N6 *U1=0 *U2=0 *U3=1 *V2=0 *X0=2048 *X1=2 S00=001 S02=043 S03=013 S04=010 S05=008 S06=002 S07=060 S08=002 S09=006 S10=007 S11=070 S12=050 S13=000 S15=000 S19=000 S21=010 S22=017 S23=019 S24=150 S25=005 S26=001 S27=000 S28=008 S29=020 S31=000 S32=009 S33=000 S34=000 S35=000 S36=000 S37=000 S38=000 S39=011 S40=000 S41=000 S42=126 S43=200 S44=015 S46=255 S47=032 S48=000 S49=016 S50=100 S51=000 S52=005 S53=000 S54=064 S55=000 S56=000 S57=000 S58=000 S60=000 S61=000 S62=000 S63=000 S64=000 S65=000 S66=000 S67=001 S68=000 S69=000 S70=000 S71=001 S72=000 S73=001 S74=000 S75=000 S76=000 S77=000 S78=000 How can we set the rack so that outbound calls are not strangled like this? All efforts so far tell us we cannot set the DTE rate on a packetbus device. Thanks. DCP

Can someone email me some of the pitfalls when ordering CHT1 service, and also some of the disadvantages over PRI (such as robbed bit signalling etc). Thanks! 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 -----------------------------/

Does any have a pmwho or tcwho that will work with a TC? Also does any have the the steps to make MRTG work with a TC? -- Matthew Opoka Systems Admin Mississippi Internet Services, Inc. http://www.magnolia.net Voice: 601.661.0081 Fax: 601.634.1952

Well, I've had my new TC up for less than a week now and twice I've run into this problem. I'll dial one of the lead numbers on a channelised T1 and I'll get nothing but "dead air". The weird part about it is, sometimes if I just wait on the line listening to silence for about 30 seconds to a minute after placing the call, one of the modems will finally pick up. Other times I will call and get dead air and it won't pick up until after I hang up. I know this because the times I don't get any carrier, immediately after I put down the receiver, one of the modem lights on a quad card picks up. I thought this to be coincedence at first, but I've been able to repeat this several times already. The CT1 is provisioned as such. ESF, B8ZS, E&M II Immediate, DTMF The dual T1 card has a HW REV of 4.0.0 and a SW REV of 4.2.1. The netserver card is HW REV 7.0.0 and SW REV of 3.6.28. And finally, the NMC is HW REV 6.0 and SW REV of 5.1.8. The only way I've found to fix this problem so far is to do a total power cycle on the whole chassis. This is definitely not something I like doing and knocking off all those users already on definitely doesn't make for happy customers. I have already tested the lines with the phone company and everything looks OK on their end. The calls are hitting the CT1's and they are seeing a seizure, but the modems just aren't picking up. Anyhow, has anyone else run into this problem? If so, any ideas on what might be causing this behaviour? And even more so, anyone know how to fix it? ============================================================================= 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. =============================================================================

Does anyone know what the acceptable internal operating temperatures are for the usr-tcs? I have two that were just set up in a remote location. They aren't handling calls yet, but when I look at the SNMP stats, I see they are running about 100-105 deg. F. Thanks, Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

I can't imagine a tcwho should be too hard, if anyone wants to do it via perl let me know and I have a lovely little perl module that will open a telnet socket, send & expect strings, etc. As for mrtg, here's a little config piece that will get you traffic statistics for the lan portion of your rack. I'd do more but the folks who came before me were (and still are) rather smart and set our radius to log to a database so we grab user stats via that :) Target[tc00.austin.1]: 1:<community>@tc00.austin MaxBytes[tc00.austin.1]: 1250000 Options[tc00.austin.1]: growright, bits WithPeak[tc00.austin.1]: ymw Directory[tc00.austin.1]: total-rate Title[tc00.austin.1]: (tc00.austin.texas.net): U.S.RoboticsTCP ARP PageTop[tc00.austin.1]: <H1>Traffic Analysis for U.S.RoboticsTCP ARP </H1> <TABLE> <TR><TD>System:</TD><TD> in </TD></TR> <TR><TD>Maintainer:</TD><TD></TD></TR> <TR><TD>Interface:</TD><TD>U.S.RoboticsTCP ARP (1)</TD></TR> <TR><TD>IP:</TD><TD>tc00.austin.texas.net</TD></TR> <TR><TD>Max Speed:</TD> <TD>1250.0 kBytes/s (ethernetCsmacd)</TD></TR> </TABLE> On Sat, Jan 17, 1998 at 10:47:19PM -0600, Matthew Opoka made some electrons appear in the following form: > Does any have a pmwho or tcwho that will work with a TC? > Also does any have the the steps to make MRTG work with a TC? > > -- > Matthew Opoka > Systems Admin > Mississippi Internet Services, Inc. > http://www.magnolia.net > Voice: 601.661.0081 > Fax: 601.634.1952 > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. --

: Does anyone know what the acceptable internal operating temperatures : are for the usr-tcs? I have two that were just set up in a remote : location. They aren't handling calls yet, but when I look at the SNMP : stats, I see they are running about 100-105 deg. F. I believe that the chassis installation manual puts the maximum operating temperature at 40 degrees centigrade, or 104 degrees Faherenheit. Others on the list have said that they've run TC units several degrees warmer than that with no problems; somebody said that theirs started locking up at 125'F, and a fellow from Japan mentioned something about a fire. At your temperature, it might be worth installing another fan tray (don't overpay -- lots of companies make them) or having an air-conditioning duct run. --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190

Did anyone step up to be the FAQ coordinator? Thanks.

how can you add more than a primary and 1 secondary radius server in the netserver? the docs for the new version of the nmc code mentions up to 8 radius servers AND some traps and even an option to let people on if everything has gone to hell and none of the radius servers is reachable. how do i do this in the netservers? matthew

> Our dialin lines seem to be working fine. Plenty of 33.6/31.2 connects all > around and good throughput. However, dialouts are severely limited. When a > user telnets to a device for dialout, connects are limited to 19.2k. This > seems to be directly related to the modem's reported DTE rate (19.2). For > instance, an ATI5 returns: > > USRobotics Digital Quad NVRAM Settings... > Copyright, 1988-97, U.S. Robotics. All rights reserved. > > DIAL=TONE B0 F1 X7 > BAUD=19200 PARITY=N WORDLEN=8 > . > . > . That's the speed the modem has written in NVRAM. It can be changed from terminal using AT&W, or from management console: Programmed Settings / DTE Interface Settings / Default Data Rate (&Bn) / ... Kamil Kukura UNICOM, Usti nad Labem Czech Republic

> Does anyone know what the acceptable internal operating temperatures are > for the usr-tcs? I have two that were just set up in a remote location. > They aren't handling calls yet, but when I look at the SNMP stats, I see > they are running about 100-105 deg. F. I don't know what is acceptable, but I've had some running at over 115 degrees for a couple of months with no problems. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

> Well, I've had my new TC up for less than a week now and twice I've > run into this problem. I'll dial one of the lead numbers on a channelised > T1 and I'll get nothing but "dead air". > > Anyhow, has anyone else run into this problem? If so, any ideas on what > might be causing this behaviour? And even more so, anyone know how to > fix it? I've seen this quite a few times. It usually relates to the number of digits that get passed from the phone company, and how many you are expecting on your end. Depending on the switch type, the modem will sit and wait for the DNIS digits before starting its negotiations. Try setting the DNIS digits to 0 and see what happens, or at least confirm with the phone company what they are sending and configure to that. Although, I've had Ameritech techs swear they were sending me 4 digits when they were only sending 1. You can often confirm this by doing a performance monitor via TCM on the modem that goes off-hook, and see what DNIS digits were sent. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

i just checked and ours average between 27c and 30c depending on whether they are in the top of the rack or the bottom. we did a bit of experiementation: first all of our chassis are in apc netshelter racks. second putting 1u blank spacers between each unit really helps. third putting nothing above the last rack for at least 10 inches really helps. we found by putting a router or switch just 1 or 2 u above the top tc in the rack made a big difference in the temperature. we were able to save 4 or 5 degrees celsius by putting 4 5.25 fans on top of the rack drawing air out. also, closing the door on the rack and filling ALL the blank spaces really does help cool it all down. matthew Curt Shambeau wrote: > > > Does anyone know what the acceptable internal operating temperatures are > > for the usr-tcs? I have two that were just set up in a remote location. > > They aren't handling calls yet, but when I look at the SNMP stats, I see > > they are running about 100-105 deg. F. > > I don't know what is acceptable, but I've had some running at over 115 > degrees for a couple of months with no problems. > > -------------------------------------------------------------------------- > | 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. -- the spa! online services - www.the-spa.com - info@the-spa.com 654b new ludlow road, south hadley, ma 01075 - 413 539-9818 providing online services since 1989!

We have a modem that picks up, I think it may be a modem on an HDM, but I can't be sure, that emits a singel monotone tone. Very flat sounding, and users modems can't connect to it. Does anyone know what this could be? 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 -----------------------------/

On Sun, 18 Jan 1998, Mark R. Lindsey wrote: > Did anyone step up to be the FAQ coordinator? Granted, the FAQ pages on totalservice.usr.com are a bit out of date and not very exhaustive, AND the list could use one. However, due to the nature of these beasts, that FAQ will be equally beasty. You have modems, four different kinds (analog, digital, analog/digital, HDM) without mentioning the dual 14.4 cards. You have everything to do with span provisioning, TCM, MPIP, PPTP, Netserver issues, ISDN, routing, RADIUS and everything that goes with authentication (CHAP, PAP, providing shell access, preventing duplicate logins), SNMP via the NMC, and any number of interoperability issues. It's the place where the rubber meets the road in the dialup 'net access business, so compiling a single FAQ on everything discussed here is enormous in scope. I'd like to see someone try it. :) That said, I might take on a specific piece of the pie. -CH

At 08:46 AM 1/18/98 -0500, Mark R. Lindsey wrote: >: Does anyone know what the acceptable internal operating temperatures >: are for the usr-tcs? I have two that were just set up in a remote >: location. They aren't handling calls yet, but when I look at the SNMP >: stats, I see they are running about 100-105 deg. F. > >I believe that the chassis installation manual puts the maximum >operating temperature at 40 degrees centigrade, or 104 degrees >Faherenheit. Others on the list have said that they've run TC units >several degrees warmer than that with no problems; somebody said that >theirs started locking up at 125'F, and a fellow from Japan mentioned >something about a fire. Peter may have been exagerating just a little when he said "burst into flames". :-) Actually what happened was the fans stopped on a chassis at a remote access point...too bad I didn't have the snmp chassis temperature being logged somewhere or I might have been able to have given the definitive answer to the operating temperature question... 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...

I'm using the pmwho with both LV and Netservers and have had no problems. >Date: Sun, 18 Jan 1998 08:43:26 -0600 >From: Brett Hawn <blh@texas.net> >To: usr-tc@lists.xmission.com >Subject: Re: (usr-tc) utils for TC >X-Mailer: Mutt 0.88 >Sender: owner-usr-tc@lists.xmission.com >Reply-To: usr-tc@lists.xmission.com > >I can't imagine a tcwho should be too hard, if anyone wants to do it via >perl let me know and I have a lovely little perl module that will open a >telnet socket, send & expect strings, etc. > >As for mrtg, here's a little config piece that will get you traffic >statistics for the lan portion of your rack. I'd do more but the folks who >came before me were (and still are) rather smart and set our radius to log >to a database so we grab user stats via that :) > >Target[tc00.austin.1]: 1:<community>@tc00.austin >MaxBytes[tc00.austin.1]: 1250000 >Options[tc00.austin.1]: growright, bits >WithPeak[tc00.austin.1]: ymw >Directory[tc00.austin.1]: total-rate >Title[tc00.austin.1]: (tc00.austin.texas.net): U.S.RoboticsTCP ARP >PageTop[tc00.austin.1]: <H1>Traffic Analysis for U.S.RoboticsTCP ARP > </H1> > <TABLE> > <TR><TD>System:</TD><TD> in </TD></TR> > <TR><TD>Maintainer:</TD><TD></TD></TR> > <TR><TD>Interface:</TD><TD>U.S.RoboticsTCP ARP (1)</TD></TR> > <TR><TD>IP:</TD><TD>tc00.austin.texas.net</TD></TR> > <TR><TD>Max Speed:</TD> > <TD>1250.0 kBytes/s (ethernetCsmacd)</TD></TR> > </TABLE> > > >On Sat, Jan 17, 1998 at 10:47:19PM -0600, Matthew Opoka made some electrons appear in the following form: >> Does any have a pmwho or tcwho that will work with a TC? >> Also does any have the the steps to make MRTG work with a TC? >> >> -- >> Matthew Opoka >> Systems Admin >> Mississippi Internet Services, Inc. >> http://www.magnolia.net >> Voice: 601.661.0081 >> Fax: 601.634.1952 >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the 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. > >

So I went out and got some NVRAM, same part # as in my newer tc racks, and thought "golly, this should be a snap".. it isn't. Has anyone upgraded their older NMC cards from 2mb to 4mb FLASH and had it work? If so what did you do? --

At 08:43 AM 1/18/98 -0600, you wrote: >I can't imagine a tcwho should be too hard, if anyone wants to do it via >perl let me know and I have a lovely little perl module that will open a >telnet socket, send & expect strings, etc. Already done.. I sent it to this list a while ago.. I will send it again if anyone else needs it.. It is perl based and easy to modify.. -M

At 02:21 PM 1/18/98 -0500, you wrote: > > how can you add more than a primary and 1 secondary radius server in the >netserver? > > the docs for the new version of the nmc code mentions up to 8 radius >servers AND some traps and even an option to let people on if everything >has gone to hell and none of the radius servers is reachable. > > how do i do this in the netservers? > You dont.. That feature was implemented for customers who dont use Netserver cards.. (I dont know why, so dont ask).. At this time you are limited to 2 RADIUS servers on the netserver.. -M

This is a multi-part message in MIME format. --------------327B0D37C983F2ED495CBD77 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > Curt Shambeau wrote: > > Does anyone know what the acceptable internal operating temperatures are > > for the usr-tcs? I have two that were just set up in a remote location. > > They aren't handling calls yet, but when I look at the SNMP stats, I see > > they are running about 100-105 deg. F. > > >> I don't know what is acceptable, but I've had some running at over 115 > >> degrees for a couple of months with no problems. Normally my two chassis run anywhere from 28 to 32 degrees Celcius. However, I had a power supply fail and the temp went up to app 40 degrees Celcius. I was notified in two ways of the problem. In the alarm server I got a message that the temp was up around 40 and the hub status light went red on the management card. (This was after I replaced the bad power supply, by the way.) Eventually the hub cooled down and the status light went back to green. Jim Pergolizzi Autodesk Network Engineer > > > -------------------------------------------------------------------------- > | 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. --------------327B0D37C983F2ED495CBD77 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Pergolizzi, Jim Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Jim Pergolizzi n: Pergolizzi;Jim org: Autodesk, Inc. adr: 3900 Civic Center Dr.;;;San Rafael;CA;94903;USA email;internet: jim.pergolizzi@autodesk.com title: Network Engineer tel;work: 415-507-8897 tel;fax: 415-507-8395 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------327B0D37C983F2ED495CBD77--

You have to run pcsdl to flash the NMC code onto the new chips. -CH On Mon, 19 Jan 1998, Brett Hawn wrote: > So I went out and got some NVRAM, same part # as in my newer tc racks, and > thought "golly, this should be a snap".. it isn't. Has anyone upgraded their > older NMC cards from 2mb to 4mb FLASH and had it work? If so what did you > do? > > -- > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

On Fri, 16 Jan 1998, MegaZone wrote: > This is COMPLETE BULLSHIT. The PM3s that I had were okay, but none of our old legacy customers could connect into them. New modems are okay, but we've got a lot of customers that haven't upgraded in two years and they were unable to connect. If it hadn't been for that, we'd probably still have some PM3s. Since then, we've bought another TNT and six MAX 4048s. No one at Livingston or anywhere else was able to solve this problem. PM3s and Ascend Pipelines don't really love each other either, and since we've got so many pipelines out there, the PM3 was really a bad mistake for us to make. For someone starting up a new ISP, or for someone who's migrating away from modems on the wall, the PM3 vs. Ascend issue is one to consider, but in my humble opinion, Ascend's still got the upper hand overall. When I upgraded my MAX 4048s to K56flex modems, the majority of our customers didn't know it until I told them. When I put the PM3s online, our tech support desk lit up like a christmas tree. The bottom line is, everyone has preferences and everyone has their own experiences. Go with whatever you feel most comfortable with. I personally don't think either product, PM3 or Ascend, is any better than the other. Each one does the same job as well as the other, only the user interface is different. Regards, Jason. -- / / o Jason Buchanan o Digistar Microsystems / OnTheNet, LLC / jsb@digistar.com o http://www.digistar.com/~jsb/

Problem here is that I couldn't even get the bloody thing to answer on a hardwire connection to a wyse 100. Checked the cable on another nmc card and the netserver card of the tc rack in question, both answered. On Mon, Jan 19, 1998 at 09:52:38AM -0600, Charles Hill made some electrons appear in the following form: > > You have to run pcsdl to flash the NMC code onto the new chips. -CH > > On Mon, 19 Jan 1998, Brett Hawn wrote: > > > So I went out and got some NVRAM, same part # as in my newer tc racks, and > > thought "golly, this should be a snap".. it isn't. Has anyone upgraded their > > older NMC cards from 2mb to 4mb FLASH and had it work? If so what did you > > do? > > > > -- > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. --

On Mon, 19 Jan 1998 pm3@digistar.com wrote: > that haven't upgraded in two years and they were unable to connect. If it > hadn't been for that, we'd probably still have some PM3s. Since then, > we've bought another TNT and six MAX 4048s. No one at Livingston or > anywhere else was able to solve this problem. PM3s and Ascend Pipelines The overwhelming problems we had with the maxes forced US to no longer use them. We had some problems with modems and the PM 3's at first, but we have all of them resolved. We also looked at the falling stock prices, the press releases, etc... things that can affect future production. > don't really love each other either, and since we've got so many pipelines > out there, the PM3 was really a bad mistake for us to make. We still have a few pipes dialing into pm 3's and there is no problem. We even get compression. Id replace em with OR-U's if I needed to and if OR-U's had pots at the time. > For someone starting up a new ISP, or for someone who's migrating away > from modems on the wall, the PM3 vs. Ascend issue is one to consider, but > in my humble opinion, Ascend's still got the upper hand overall. If you had the problems we had (memory leaks, session overlapping, overheating) you would NEVER want to use the 4000. > When I upgraded my MAX 4048s to K56flex modems, the majority of our > customers didn't know it until I told them. When I put the PM3s online, > our tech support desk lit up like a christmas tree. That is interesting and perhaps valid. That seems to be one thing that is going in Ascend's direction and hurting the PM 3 are the connection issues. As mentioned before, we have them all resolved, but Im not really a good case, as most of our customers are dedicated 24/7 OR they use the modems I sell to them. > The bottom line is, everyone has preferences and everyone has their own > experiences. Go with whatever you feel most comfortable with. I > personally don't think either product, PM3 or Ascend, is any better than > the other. Each one does the same job as well as the other, only the user > interface is different. My decision was based on a lot of things, service, support, quality of hardware, stock movement, market movement, etc... and Im the kind of person that uses the best tool for the job. And of course, I am going to tout the product that we most often like to sell and service, AND USE. My primary issue with the previous posting was the comment that the PM 3 was cheaply made. They are definitely NOT cheaply made. ~~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~ Jake Messinger ph:713-772-6690 Livingston Var AMS, Inc. fx:713-774-3498 Medical Billing 8300 Bissonnet #400 jake@ams.com Internet Services Houston, Texas 77074 www.ams.com/~jake Business Management ~~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~

On Mon, 19 Jan 1998 pm3@digistar.com wrote: > that haven't upgraded in two years and they were unable to connect. If it > hadn't been for that, we'd probably still have some PM3s. Since then, > we've bought another TNT and six MAX 4048s. No one at Livingston or > anywhere else was able to solve this problem. PM3s and Ascend Pipelines The overwhelming problems we had with the maxes forced US to no longer use them. We had some problems with modems and the PM 3's at first, but we have all of them resolved. We also looked at the falling stock prices, the press releases, etc... things that can affect future production. > don't really love each other either, and since we've got so many pipelines > out there, the PM3 was really a bad mistake for us to make. We still have a few pipes dialing into pm 3's and there is no problem. We even get compression. Id replace em with OR-U's if I needed to and if OR-U's had pots at the time. > For someone starting up a new ISP, or for someone who's migrating away > from modems on the wall, the PM3 vs. Ascend issue is one to consider, but > in my humble opinion, Ascend's still got the upper hand overall. If you had the problems we had (memory leaks, session overlapping, overheating) you would NEVER want to use the 4000. > When I upgraded my MAX 4048s to K56flex modems, the majority of our > customers didn't know it until I told them. When I put the PM3s online, > our tech support desk lit up like a christmas tree. That is interesting and perhaps valid. That seems to be one thing that is going in Ascend's direction and hurting the PM 3 are the connection issues. As mentioned before, we have them all resolved, but Im not really a good case, as most of our customers are dedicated 24/7 OR they use the modems I sell to them. > The bottom line is, everyone has preferences and everyone has their own > experiences. Go with whatever you feel most comfortable with. I > personally don't think either product, PM3 or Ascend, is any better than > the other. Each one does the same job as well as the other, only the user > interface is different. My decision was based on a lot of things, service, support, quality of hardware, stock movement, market movement, etc... and Im the kind of person that uses the best tool for the job. And of course, I am going to tout the product that we most often like to sell and service, AND USE. My primary issue with the previous posting was the comment that the PM 3 was cheaply made. They are definitely NOT cheaply made. ~~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~ Jake Messinger ph:713-772-6690 Livingston Var AMS, Inc. fx:713-774-3498 Medical Billing 8300 Bissonnet #400 jake@ams.com Internet Services Houston, Texas 77074 www.ams.com/~jake Business Management ~~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~

I have installed the security and accounting database on my Sparcstation running Solaris 2.5.1 with the Postgres database. Fact is, I can't access it with ODBC or SQL server, so I can manage it remote on an NT with M$ Access. Thanks for any pointers, Robert von Bismarck Petrel Communication S.A.

We also have been running into this problem on just one specific channel. Previously had thought it was a problem with the modem on that channel, but switching the Quad modem with one on another slot proved otherwise. Have called the phone company, but without providing the specific problem and a detailed solution, they are powerless. My question is this: why would the telco send different DNIS digits for one channel than another on the same T-span? On Sun, 18 Jan 1998, Curt Shambeau wrote: > I've seen this quite a few times. It usually relates to the number of > digits that get passed from the phone company, and how many you are > expecting on your end. > > Depending on the switch type, the modem will sit and wait for the DNIS > digits before starting its negotiations. Try setting the DNIS digits to 0 > and see what happens, or at least confirm with the phone company what they > are sending and configure to that. > > Although, I've had Ameritech techs swear they were sending me 4 digits > when they were only sending 1. You can often confirm this by doing a > performance monitor via TCM on the modem that goes off-hook, and see what > DNIS digits were sent.

Brian said once upon a time: > >We have a modem that picks up, I think it may be a modem on an HDM, but I >can't be sure, that emits a singel monotone tone. Very flat sounding, and >users modems can't connect to it. > >Does anyone know what this could be? I've had this happen on Quads, but not on HDMs yet (to my knowledge). The tone mode is one of the tone-test modes. You essentially need to issue a disconnect on all your channels and reset the modems to fix it in the quick fashion. Watching the calls on via the performance meter can fix it in the slow fashion, just look for the channel/modem that isn't answering anything and issue a disconnect/reset on that.

> We also have been running into this problem on just one specific channel. > Previously had thought it was a problem with the modem on that channel, > but switching the Quad modem with one on another slot proved otherwise. > Have called the phone company, but without providing the specific problem > and a detailed solution, they are powerless. My question is this: why > would the telco send different DNIS digits for one channel than another on > the same T-span? They shouldn't. I wasn't aware that the problem was happening on just one channel of the T1. The number of digits should be based on the T1, not on individual channels. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

This rant is made by someone who is obviously confused. The Ascends are good for two reasons; heavy doorstops, and space heaters. On Mon, 19 Jan 1998 pm3@digistar.com wrote: > For someone starting up a new ISP, or for someone who's migrating away > from modems on the wall, the PM3 vs. Ascend issue is one to consider, but > in my humble opinion, Ascend's still got the upper hand overall. >

I have seen the same situation in a rack or two of my own. A couple of channels being dead, etc... Being the Telco we checked the Ts and what not and they are perfectly fine. I even ran the Ts on a non-USR rack and all channels worked out fine. The interesting thing is that we filter out DNIS digits out at the switch. (5ESS) We have no real need for them right now. I have the T NACs set for no addressing, so that shouldn't be a problem. I have even swapped out the Quad card, but no success. Odd... -Chris > > We also have been running into this problem on just one specific channel. > Previously had thought it was a problem with the modem on that channel, > but switching the Quad modem with one on another slot proved otherwise. > Have called the phone company, but without providing the specific problem > and a detailed solution, they are powerless. My question is this: why > would the telco send different DNIS digits for one channel than another on > the same T-span? > > On Sun, 18 Jan 1998, Curt Shambeau wrote: > > I've seen this quite a few times. It usually relates to the number of > > digits that get passed from the phone company, and how many you are > > expecting on your end. > > > > Depending on the switch type, the modem will sit and wait for the DNIS > > digits before starting its negotiations. Try setting the DNIS digits to 0 > > and see what happens, or at least confirm with the phone company what they > > are sending and configure to that. > > > > Although, I've had Ameritech techs swear they were sending me 4 digits > > when they were only sending 1. You can often confirm this by doing a > > performance monitor via TCM on the modem that goes off-hook, and see what > > DNIS digits were sent. > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

I used the USR official flash upgrade module. The problem is they don't ship it with code on it, so you've got to PCSDL the card once you put the new flash in, otherwise it won't boot. At 09:02 AM 1/19/98 -0600, you wrote: >So I went out and got some NVRAM, same part # as in my newer tc racks, and >thought "golly, this should be a snap".. it isn't. Has anyone upgraded their >older NMC cards from 2mb to 4mb FLASH and had it work? If so what did you >do? > >-- > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > >

We had a problem a few weeks ago, when the telco configured a new trunk group as DT Wink. The USR's out of the box expect MF signalling instead. You may want to set the Dual T1 card to NO signalling, see if they answer, then debug it from there. At 12:48 PM 1/19/98 -0500, you wrote: >I have seen the same situation in a rack or two of my own. A couple of >channels being dead, etc... Being the Telco we checked the Ts and what not >and they are perfectly fine. I even ran the Ts on a non-USR rack and all >channels worked out fine. > >The interesting thing is that we filter out DNIS digits out at the >switch. (5ESS) We have no real need for them right now. I have the T >NACs set for no addressing, so that shouldn't be a problem. I have even >swapped out the Quad card, but no success. Odd... > >-Chris > > >> >> We also have been running into this problem on just one specific channel. >> Previously had thought it was a problem with the modem on that channel, >> but switching the Quad modem with one on another slot proved otherwise. >> Have called the phone company, but without providing the specific problem >> and a detailed solution, they are powerless. My question is this: why >> would the telco send different DNIS digits for one channel than another on >> the same T-span? >> >> On Sun, 18 Jan 1998, Curt Shambeau wrote: >> > I've seen this quite a few times. It usually relates to the number of >> > digits that get passed from the phone company, and how many you are >> > expecting on your end. >> > >> > Depending on the switch type, the modem will sit and wait for the DNIS >> > digits before starting its negotiations. Try setting the DNIS digits to 0 >> > and see what happens, or at least confirm with the phone company what they >> > are sending and configure to that. >> > >> > Although, I've had Ameritech techs swear they were sending me 4 digits >> > when they were only sending 1. You can often confirm this by doing a >> > performance monitor via TCM on the modem that goes off-hook, and see what >> > DNIS digits were sent. >> >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the 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. > >

Unfortunatly, USR changed the display format on the latest version of the Netserver code. It now spaces the session status over by one space to the right. Causes our version of PMWHO to display each and every port, not just the active ones. I'm going to have to dig through the code and make it USR aware. At 08:13 AM 1/19/98 -0600, you wrote: > >I'm using the pmwho with both LV and Netservers and have had no problems. > > > >>Date: Sun, 18 Jan 1998 08:43:26 -0600 >>From: Brett Hawn <blh@texas.net> >>To: usr-tc@lists.xmission.com >>Subject: Re: (usr-tc) utils for TC >>X-Mailer: Mutt 0.88 >>Sender: owner-usr-tc@lists.xmission.com >>Reply-To: usr-tc@lists.xmission.com >> >>I can't imagine a tcwho should be too hard, if anyone wants to do it via >>perl let me know and I have a lovely little perl module that will open a >>telnet socket, send & expect strings, etc. >> >>As for mrtg, here's a little config piece that will get you traffic >>statistics for the lan portion of your rack. I'd do more but the folks who >>came before me were (and still are) rather smart and set our radius to log >>to a database so we grab user stats via that :) >> >>Target[tc00.austin.1]: 1:<community>@tc00.austin >>MaxBytes[tc00.austin.1]: 1250000 >>Options[tc00.austin.1]: growright, bits >>WithPeak[tc00.austin.1]: ymw >>Directory[tc00.austin.1]: total-rate >>Title[tc00.austin.1]: (tc00.austin.texas.net): U.S.RoboticsTCP ARP >>PageTop[tc00.austin.1]: <H1>Traffic Analysis for U.S.RoboticsTCP ARP >> </H1> >> <TABLE> >> <TR><TD>System:</TD><TD> in </TD></TR> >> <TR><TD>Maintainer:</TD><TD></TD></TR> >> <TR><TD>Interface:</TD><TD>U.S.RoboticsTCP ARP (1)</TD></TR> >> <TR><TD>IP:</TD><TD>tc00.austin.texas.net</TD></TR> >> <TR><TD>Max Speed:</TD> >> <TD>1250.0 kBytes/s (ethernetCsmacd)</TD></TR> >> </TABLE> >> >> >>On Sat, Jan 17, 1998 at 10:47:19PM -0600, Matthew Opoka made some >electrons appear in the following form: >>> Does any have a pmwho or tcwho that will work with a TC? >>> Also does any have the the steps to make MRTG work with a TC? >>> >>> -- >>> Matthew Opoka >>> Systems Admin >>> Mississippi Internet Services, Inc. >>> http://www.magnolia.net >>> Voice: 601.661.0081 >>> Fax: 601.634.1952 >>> >>> >>> - >>> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >>> with "unsubscribe usr-tc" in the body of the 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. > >

At 10:10 AM 1/19/98 -0600, you wrote: >The PM3s that I had were okay, but none of our old legacy customers could >connect into them. New modems are okay, but we've got a lot of customers >that haven't upgraded in two years and they were unable to connect. If it >hadn't been for that, we'd probably still have some PM3s. Since then, >we've bought another TNT and six MAX 4048s. No one at Livingston or >anywhere else was able to solve this problem. PM3s and Ascend Pipelines >don't really love each other either, and since we've got so many pipelines >out there, the PM3 was really a bad mistake for us to make. > Thats odd. We have very good luck with our PM3's. Almost all of our early ISDN customers used Pipe 50's, and they work great with the PM3. We even use a Pipe 50 to connect our admin offices into our network, via a PM3. > >

I have installed the security and accounting database on my Sparcstation running Solaris 2.5.1 with the Postgres database. Fact is, I can't access it with ODBC or SQL server, so I can manage it remote on an NT with M$ Access. Thanks for any pointers, Robert von Bismarck Petrel Communication S.A

: > Did anyone step up to be the FAQ coordinator? : : Granted, the FAQ pages on totalservice.usr.com are a bit out of date and : not very exhaustive, AND the list could use one. However, due to the : nature of these beasts, that FAQ will be equally beasty. I don't mean to compile a linuxesque Howto or some sort of complete documentation set: I mean a list of questions that come up regularly on this list. There really aren't that many, if you think about it. Most questions are fairly uncommon; the list archives are there for such purposes. A FAQ would cover the stuff that comes up again and again. Mark

While we're on the topic, is there a searchable archive of this list? I know that's one that would be in the FAQ, if there were one... Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Mon, 19 Jan 1998, Mark R. Lindsey wrote: > Date: Mon, 19 Jan 1998 19:02:11 -0500 > From: "Mark R. Lindsey" <mark@vielle.datasys.net> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) FAQ coordinator? > > : > Did anyone step up to be the FAQ coordinator? > : > : Granted, the FAQ pages on totalservice.usr.com are a bit out of date and > : not very exhaustive, AND the list could use one. However, due to the > : nature of these beasts, that FAQ will be equally beasty. > > I don't mean to compile a linuxesque Howto or some sort of complete > documentation set: I mean a list of questions that come up regularly on > this list. There really aren't that many, if you think about it. Most > questions are fairly uncommon; the list archives are there for such > purposes. A FAQ would cover the stuff that comes up again and again. > > Mark > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Mmm.. here's one for ya right off the top that is nowhere to be found on totalservice (according to their search engines). How to use PCSDL properly, there's nothing more annoying that a really pathetic hook a computer up to dialup equipment and load some crud via xmodem or whatever protocol used. I'm VERY VERY VERY dissapointed in USR/3Com in that they sent me 2 new NMC cards with nothing on the flash. If you're going to do this at least have the decency to provide a _REAL_ tool for dealing with this. Dragging a laptop into my noc just to deal with your incompetence is pretty annoying. On Mon, Jan 19, 1998 at 07:02:11PM -0500, Mark R. Lindsey made some electrons appear in the following form: > : > Did anyone step up to be the FAQ coordinator? > : > : Granted, the FAQ pages on totalservice.usr.com are a bit out of date and > : not very exhaustive, AND the list could use one. However, due to the > : nature of these beasts, that FAQ will be equally beasty. > > I don't mean to compile a linuxesque Howto or some sort of complete > documentation set: I mean a list of questions that come up regularly on > this list. There really aren't that many, if you think about it. Most > questions are fairly uncommon; the list archives are there for such > purposes. A FAQ would cover the stuff that comes up again and again. > > Mark > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. --

> Thought i'd just put in a good word for the USR/3com box here: > We're an ISP who is in the latter category, ie we've just moved away from > the analogue modem-bank nightmare. We've had zero problems with the TC so > far, our users are getting good connect speeds, good throughput. We've had nothing but trouble with the TC stuff. They started out ok, but then we started to get more and more unexplained disconnects. After that, the hardware started to fail left and right for no apparent reason. Then, a 6 WEEK turnaround on getting bad components repaired, and we're only a FEW MILES from their offices! And the clincher was that when we did finally get the components back, they didn't drop any calls, but didn't even power up!!!! We're trying out a Bay box for x2 and so far so good.

> Thought i'd just put in a good word for the USR/3com box here: > We're an ISP who is in the latter category, ie we've just moved away from > the analogue modem-bank nightmare. We've had zero problems with the TC so > far, our users are getting good connect speeds, good throughput. We've had nothing but trouble with the TC stuff. They started out ok, but then we started to get more and more unexplained disconnects. After that, the hardware started to fail left and right for no apparent reason. Then, a 6 WEEK turnaround on getting bad components repaired, and we're only a FEW MILES from their offices! And the clincher was that when we did finally get the components back, they didn't drop any calls, but didn't even power up!!!! We're trying out a Bay box for x2 and so far so good.

> Well all i can say is i hope we don't get any of your problems. We've had > no disconnects reported. Obviously our two completely different > experiences do raise some QA concerns about the USR kit. :( We have 3 chassis, the oldest about 6 months old. Only the 2 month old one hasn't failed seriously yet. The other two have lost their configs for no apparent reason, and then the power supply problems started. *SIGH*

I am getting ready to purchase the ISPTRACK ( http://www.cyberacs.com ) software, and courious how you get the file out of the Radius Software for processing... Any advice would be appreaciated. 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

Clayton Zekelman said: : Unfortunatly, USR changed the display format on the latest version of : the Netserver code. It now spaces the session status over by one : space to the right. Causes our version of PMWHO to display each and : every port, not just the active ones. I'm going to have to dig through : the code and make it USR aware. Isn't this the EXACT sort of problem that caused a big group of people to get together and create the Simple Network Management Protocol, SNMP? Is it too much to ask USR/3Com to make the information available by `show sessions' available via SNMP, too? (Who do I have to call and beg?) --- Mark R. Lindsey, mark@datasys.net Internet Engineer, DSS Online Voice: +1 912 241 0607; Fax: +1 912 241 0190

Speaking of the latest version of code, has anyone else noticed that the NMC no longer reports the DNIS or Caller ID (ANI) of an incoming call? One of our boxes is on the old code and it still logs them, but the other boxes used to, and don't anymore since the so-called "upgrade". (Though the upgrade did help quite a bit with the problems we were having with MPIP) The DNIS and ANI (Caller ID)used to be viewable through TCM, SNMP and sent through RADIUS, but not since we flashed to TCS 3.0 code. We have dual PRI cards, and a standard 48 modem setup. Any suggestions or shared experiences? Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- > >Clayton Zekelman said: >: Unfortunatly, USR changed the display format on the latest version of >: the Netserver code. It now spaces the session status over by one >: space to the right. Causes our version of PMWHO to display each and >: every port, not just the active ones. I'm going to have to dig through >: the code and make it USR aware. > >Isn't this the EXACT sort of problem that caused a big group of people to >get together and create the Simple Network Management Protocol, SNMP? > >Is it too much to ask USR/3Com to make the information available by >`show sessions' available via SNMP, too? > >(Who do I have to call and beg?) > > > >--- >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. >

-> Speaking of the latest version of code, has anyone else noticed that the NMC -> no longer reports the DNIS or Caller ID (ANI) of an incoming call? One of -> our boxes is on the old code and it still logs them, but the other boxes -> used to, and don't anymore since the so-called "upgrade". (Though the -> upgrade did help quite a bit with the problems we were having with MPIP) The -> DNIS and ANI (Caller ID)used to be viewable through TCM, SNMP and sent -> through RADIUS, but not since we flashed to TCS 3.0 code. We have dual PRI -> cards, and a standard 48 modem setup. -> -> Any suggestions or shared experiences? Just upgraded to TCS 3.0 this past weekend and not having any problems with ANI and DNIS. Jeff Binkley ASA Network Computing

I'm not sure if it has been on the list before, so I'll ask now. Does anyone have or know of some documentation of the cause codes that the NetServer card kicks out? We have seen some strange messages from the authlog and I would like to find out what they mean. Thanks in advance! Christopher Brown ccbrown@feist.com WAN Administrator 316.337.8657 (office) ParaCom Technologies, Inc. 316.337.8607 (fax)

On Saturday 17 January 1998, at 22 h 47, the keyboard of Matthew Opoka <phantom@magnolia.net> wrote: > Does any have a pmwho or tcwho that will work with a TC? pmwho works here with a TC. Unless there is a more recent version of the code which breaks it, I'm not aware of problems. > Also does any have the the steps to make MRTG work with a TC? Nothing special for MRTG. You monitor a TC with SNMP like any other box. Do you have a more specific question?

Just was informed that UseLessWest is offering free installation on channelized T1's starting sometime in February. Have no idea if it includes PRI because the monthly rate is above outrageous in our area. Anyway, quite a saving if you can hold off an order for a month or so. 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

Thus spake Christopher Brown >I'm not sure if it has been on the list before, so I'll ask now. >Does anyone have or know of some documentation of the cause codes that the >NetServer card kicks out? We have seen some strange messages from the >authlog and I would like to find out what they mean. Go to dejanews and search in the old database for: ISDN cause codes indicates comp.dcom.isdn It'll be the article by Oliver Jones -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

On Mon, 19 Jan 1998 pm3@digistar.com wrote: > On Fri, 16 Jan 1998, MegaZone wrote: > > > This is COMPLETE BULLSHIT. > > The PM3s that I had were okay, but none of our old legacy customers could > connect into them. New modems are okay, but we've got a lot of customers > that haven't upgraded in two years and they were unable to connect. If it > hadn't been for that, we'd probably still have some PM3s. Since then, > we've bought another TNT and six MAX 4048s. No one at Livingston or > anywhere else was able to solve this problem. PM3s and Ascend Pipelines > don't really love each other either, and since we've got so many pipelines > out there, the PM3 was really a bad mistake for us to make. > > For someone starting up a new ISP, or for someone who's migrating away > from modems on the wall, the PM3 vs. Ascend issue is one to consider, but > in my humble opinion, Ascend's still got the upper hand overall. Thought i'd just put in a good word for the USR/3com box here: We're an ISP who is in the latter category, ie we've just moved away from the analogue modem-bank nightmare. We've had zero problems with the TC so far, our users are getting good connect speeds, good throughput. PM3 wasn't an option for us in NZ, they don't seem to ship here (as far as we can tell - the couple of sales enquiries we made were not answered by the Livingston people). Ascend looked ok, but overall we'd heard better things about USR so that's the way we went. And we haven't regretted it - yet. (knock on wood, etc etc) 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

On Mon, 19 Jan 1998 pm3@digistar.com wrote: > On Fri, 16 Jan 1998, MegaZone wrote: > > > This is COMPLETE BULLSHIT. > > The PM3s that I had were okay, but none of our old legacy customers could > connect into them. New modems are okay, but we've got a lot of customers > that haven't upgraded in two years and they were unable to connect. If it > hadn't been for that, we'd probably still have some PM3s. Since then, > we've bought another TNT and six MAX 4048s. No one at Livingston or > anywhere else was able to solve this problem. PM3s and Ascend Pipelines > don't really love each other either, and since we've got so many pipelines > out there, the PM3 was really a bad mistake for us to make. > > For someone starting up a new ISP, or for someone who's migrating away > from modems on the wall, the PM3 vs. Ascend issue is one to consider, but > in my humble opinion, Ascend's still got the upper hand overall. Thought i'd just put in a good word for the USR/3com box here: We're an ISP who is in the latter category, ie we've just moved away from the analogue modem-bank nightmare. We've had zero problems with the TC so far, our users are getting good connect speeds, good throughput. PM3 wasn't an option for us in NZ, they don't seem to ship here (as far as we can tell - the couple of sales enquiries we made were not answered by the Livingston people). Ascend looked ok, but overall we'd heard better things about USR so that's the way we went. And we haven't regretted it - yet. (knock on wood, etc etc) 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

On Mon, 19 Jan 1998, Gregory Gulik wrote: > They started out ok, but then we started to get more and more > unexplained disconnects. After that, the hardware started to fail > left and right for no apparent reason. > > Then, a 6 WEEK turnaround on getting bad components > repaired, and we're only a FEW MILES from their offices! > > And the clincher was that when we did finally get the components > back, they didn't drop any calls, but didn't even power up!!!! > > We're trying out a Bay box for x2 and so far so good. Well all i can say is i hope we don't get any of your problems. We've had no disconnects reported. Obviously our two completely different experiences do raise some QA concerns about the USR kit. :( 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

I have several users who are having trouble connecting and remaining connected for longer than 10 minutes. Does anyone who has seen this problem discovered a possible init string for the Global Village modems that will solve this problem? Thank you. | Cassandra M. Perkins | People usually get what's coming to | | Network Operations | them... unless it's been mailed. | | The Loop Internet Switch Co., LLC | -fortune |

I have several users who are having trouble connecting and remaining connected for longer than 10 minutes. Does anyone who has seen this problem discovered a possible init string for the Global Village modems that will solve this problem? Thank you. | Cassandra M. Perkins | People usually get what's coming to | | Network Operations | them... unless it's been mailed. | | The Loop Internet Switch Co., LLC | -fortune |

>On Saturday 17 January 1998, at 22 h 47, the keyboard of Matthew Opoka ><phantom@magnolia.net> wrote: > >> Does any have a pmwho or tcwho that will work with a TC? > >pmwho works here with a TC. Unless there is a more recent version of the >code which breaks it, I'm not aware of problems. I did this very unelegant patch (not very efficient, but it works) :) --- pmwho.c Wed Jan 14 12:47:13 1998 +++ pmwho.c.new Wed Jan 14 12:48:20 1998 @@ -520,7 +520,8 @@ fflush(stdout); } else if (strlen(line) >= IDLE_POS) { - if (strncmp(&line[IDLE_POS], "IDLE ", 5) != 0) { + if ((strstr(&line[IDLE_POS], "IDLE ") == NULL ) && + (strstr(&line[IDLE_POS], "NO-SERVICE ") == NULL ) ) { fputs(line, stdout); fflush(stdout); } This also stops it printing ports with NO-SERVICE on a PM3, if you only use TC, then you can remove the second line with a plus at the front, just swap the && on the line before for )) Save it in the same directory as pmwho.c as pmwho.patch and apply it using: patch -p0 < pmwho.patch >> Also does any have the the steps to make MRTG work with a TC? > >Nothing special for MRTG. You monitor a TC with SNMP like any other box. >Do you have a more specific question? Yes, you can set it up like any other box, but it doesn't seem to log the outgoing traffic, only the incoming. (At least it's better than the TotalSwitch, which doesn't seem to update any of it's counters in any logical manner, especially not the 100BTx ports) >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

On Tue, 20 Jan 1998, Brian wrote: > > > I was logged into a netserver of ours, and it started freaking out: > > net: Bad wanted 838 got 33 > net: Bad wanted 658 got 51 > net: Bad wanted 658 got 50 > net: Bad wanted 728 got 25 > net: Bad wanted 838 got 37 > net: Bad wanted 838 got 33 > net: Bad wanted 648 got 62 > net: Bad wanted 658 got 24 > ppp_ifsend: Bad PPP Control > net: Bad wanted 708 got 26 > net: Bad wanted 708 got 33 > net: Bad wanted 648 got 26 > net: Bad wanted 658 got 28 > net: Bad wanted 658 got 64 > net: Bad wanted 658 got 54 > net: Bad wanted 708 got 37 > net: Bad wanted 838 got 27 > net: Bad wanted 708 got 26 > ppp_ifsend: Bad PPP Control > ppp_ifsend: Bad PPP Control > net: Bad wanted 488 got 27 > > What was happening here? I have never seen anything like this before. > Normally i don't see any errors or anything ever sent to the console like > this, so i figured it must be serious. Can anyone help? > Some one has turned debug on on the NETServer - you are seeing the debug information. Do the following reset conso set debug 0 krish > > 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. >

On Tue, 20 Jan 1998, Christopher Brown wrote: > I'm not sure if it has been on the list before, so I'll ask now. > > Does anyone have or know of some documentation of the cause codes that the > NetServer card kicks out? We have seen some strange messages from the > authlog and I would like to find out what they mean. > > Thanks in advance! > > > Christopher Brown ccbrown@feist.com > WAN Administrator 316.337.8657 (office) > ParaCom Technologies, Inc. 316.337.8607 (fax) a definite FAQ question :) > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Wed, 21 Jan 1998, Brett Hawn wrote: > How about this one? I've had a ticket open with USR/3Com for over a month > and nobody has gotten back to me about what the hell this means. > > Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM This is a syslog error that you are seeing. All this means is that the NETServer is telling you that it tried to set the Async Map to the modem, and the modem did not set the same. There has been 10 pending messages to the modem and the modem has not responded back. This does not cause any problems. Anyway are you seeing this problem with HDM or Quad? krish > Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM > Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM > > > > On Tue, Jan 20, 1998 at 11:38:11PM -0600, Brian stated > > > > > > I was logged into a netserver of ours, and it started freaking out: > > > > net: Bad wanted 838 got 33 > > net: Bad wanted 658 got 51 > > net: Bad wanted 658 got 50 > > net: Bad wanted 728 got 25 > > net: Bad wanted 838 got 37 > > net: Bad wanted 838 got 33 > > net: Bad wanted 648 got 62 > > net: Bad wanted 658 got 24 > > ppp_ifsend: Bad PPP Control > > net: Bad wanted 708 got 26 > > net: Bad wanted 708 got 33 > > net: Bad wanted 648 got 26 > > net: Bad wanted 658 got 28 > > net: Bad wanted 658 got 64 > > net: Bad wanted 658 got 54 > > net: Bad wanted 708 got 37 > > net: Bad wanted 838 got 27 > > net: Bad wanted 708 got 26 > > ppp_ifsend: Bad PPP Control > > ppp_ifsend: Bad PPP Control > > net: Bad wanted 488 got 27 > > > > What was happening here? I have never seen anything like this before. > > Normally i don't see any errors or anything ever sent to the console like > > this, so i figured it must be serious. Can anyone help? > > > > > > 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. > > -- > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

I was logged into a netserver of ours, and it started freaking out: net: Bad wanted 838 got 33 net: Bad wanted 658 got 51 net: Bad wanted 658 got 50 net: Bad wanted 728 got 25 net: Bad wanted 838 got 37 net: Bad wanted 838 got 33 net: Bad wanted 648 got 62 net: Bad wanted 658 got 24 ppp_ifsend: Bad PPP Control net: Bad wanted 708 got 26 net: Bad wanted 708 got 33 net: Bad wanted 648 got 26 net: Bad wanted 658 got 28 net: Bad wanted 658 got 64 net: Bad wanted 658 got 54 net: Bad wanted 708 got 37 net: Bad wanted 838 got 27 net: Bad wanted 708 got 26 ppp_ifsend: Bad PPP Control ppp_ifsend: Bad PPP Control net: Bad wanted 488 got 27 What was happening here? I have never seen anything like this before. Normally i don't see any errors or anything ever sent to the console like this, so i figured it must be serious. Can anyone help? 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 -----------------------------/

How about this one? I've had a ticket open with USR/3Com for over a month and nobody has gotten back to me about what the hell this means. Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:18 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM Jan 18 10:11:19 tc03.sat.texas.net S31: error -19 setting remote ACCM On Tue, Jan 20, 1998 at 11:38:11PM -0600, Brian stated > > > I was logged into a netserver of ours, and it started freaking out: > > net: Bad wanted 838 got 33 > net: Bad wanted 658 got 51 > net: Bad wanted 658 got 50 > net: Bad wanted 728 got 25 > net: Bad wanted 838 got 37 > net: Bad wanted 838 got 33 > net: Bad wanted 648 got 62 > net: Bad wanted 658 got 24 > ppp_ifsend: Bad PPP Control > net: Bad wanted 708 got 26 > net: Bad wanted 708 got 33 > net: Bad wanted 648 got 26 > net: Bad wanted 658 got 28 > net: Bad wanted 658 got 64 > net: Bad wanted 658 got 54 > net: Bad wanted 708 got 37 > net: Bad wanted 838 got 27 > net: Bad wanted 708 got 26 > ppp_ifsend: Bad PPP Control > ppp_ifsend: Bad PPP Control > net: Bad wanted 488 got 27 > > What was happening here? I have never seen anything like this before. > Normally i don't see any errors or anything ever sent to the console like > this, so i figured it must be serious. Can anyone help? > > > 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. --

On Tue, 20 Jan 1998, Tatai SV Krishnan wrote: > > On Tue, 20 Jan 1998, Brian wrote: > > > > > > > I was logged into a netserver of ours, and it started freaking out: > > > > net: Bad wanted 838 got 33 > > net: Bad wanted 658 got 51 > > net: Bad wanted 658 got 50 > > net: Bad wanted 728 got 25 > > net: Bad wanted 838 got 37 > > net: Bad wanted 838 got 33 > > net: Bad wanted 648 got 62 > > net: Bad wanted 658 got 24 > > ppp_ifsend: Bad PPP Control > > net: Bad wanted 708 got 26 > > net: Bad wanted 708 got 33 > > net: Bad wanted 648 got 26 > > net: Bad wanted 658 got 28 > > net: Bad wanted 658 got 64 > > net: Bad wanted 658 got 54 > > net: Bad wanted 708 got 37 > > net: Bad wanted 838 got 27 > > net: Bad wanted 708 got 26 > > ppp_ifsend: Bad PPP Control > > ppp_ifsend: Bad PPP Control > > net: Bad wanted 488 got 27 > > > > What was happening here? I have never seen anything like this before. > > Normally i don't see any errors or anything ever sent to the console like > > this, so i figured it must be serious. Can anyone help? > > > > Some one has turned debug on on the NETServer - you are seeing the debug > information. Do the following > > reset conso > set debug 0 > > > krish > Thanks krish :) > > > > > 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. > > > /-------------------------- 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 -----------------------------/

On Mon, 19 Jan 1998, Mark R. Lindsey wrote: > Isn't this the EXACT sort of problem that caused a big group of people to > get together and create the Simple Network Management Protocol, SNMP? > > Is it too much to ask USR/3Com to make the information available by > `show sessions' available via SNMP, too? > > (Who do I have to call and beg?) 3COM/USR isn't likely to improve the SNMP on the current Netserver code. They are required to develop all new code for the Netserver this year, so they aren't likely to do anything beyond bug fixes to the current code. I'd hope the new code has better SNMP, but if the new code's as buggy as the Hiper ARC code, I don't want to try it right away. Brian

I am trying to get a customer to dial in and use NAT on there P130. I am assigning them a dynamic IP address, and it seems to be failing during PPP negotiation. On the p130 I am telling it its IP address is 192.168.100.100 as per Ascend. I think the TC hub is barfing on the IPCP stage since it doesnt like the fact the p130 wants to use 192.168.100.100. If anyone knows that NAT works with 3com/ascend p130 let me know and if you have a config for the pipline, send it over. Appreciate, 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 -----------------------------/

I installed my Netserver 16/I + and have a problem. It doesn't even try to authenticate my dial-in attempts. Here is what it gives when I dial in with Hyperterminal or something like this=A0:=20 OK atdtxxxxxxx CONNECT 115200 SPAN : Swiss Private Access Network NODE : COP-PM1 ACCESS : Restricted~#=C0!}!}!} }3}!}$}%=EA}#}%=C2#}%}%}&=B2j-G = [--SNIP-SNIP--] I followed the CLI configuration wizard, and network is okay. It must = be a simple configuration problem involving one or the other command, but I'm lost. Thank for any pointers, Robert von Bismarck Petrel Communication S.A.

On Wed, 21 Jan 1998, Charles Sprickman wrote: > Having no experience with the PRI card or setting up the Netserver for > ISDN, where is the best place to start? The most important things to set in the PRI card are switch type and ISDN-gateway slot. Usually that's enough to get it working. Since you are terminating the ISDN calls on the Netserver, the gateway slot will be the slot containing the Netserver. You'll know the PRI card is talking to the telco switch when you see the D-channel come up. If you "set all login network dialin" in the Netserver, it should take ISDN calls. If you are using RADIUS, "set all security on". . . and so on. -CH

Hello, We currently have 5 TC Hubs using channelized T1 service, and all is well. We just received an older chassis with a NMC card, Netserver card, and PRI card, and we'll be dedicating it to ISDN use. I wish to use it solely for ISDN, so there will be no modems installed in it. Having no experience with the PRI card or setting up the Netserver for ISDN, where is the best place to start? I've got an older (3.3) Netserver manual, and a ton of "hardware" manuals, but I've yet to find docs on the latest Netserver rev. Is there any type of "quick setup guide" for this? Can I do "data over voice" ISDN without the calls being routed to analog modems? Also, any suggestions on PRI provisioning are appreciated. Our telco is very accomodating and I can actually visit the CO and chat with the switch technicians once I know what to ask for... Thanks, Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com

Thank for coming back to me, here's some more info : I disabled local authentication "disable authentication local", and now it authenticates the users, but has some problems establishing a correct PPP link. I'll keep you posted if I have anything more to report. What brought me to do this, is that the debug messages from the authentication facility were saying that the internal user "default" = was successfully authenticated, whoever was calling. Regards, Robert von Bismarck Petrel Communication S.A. -----Original Message----- From: Paul_Brennan@3com.com [SMTP:Paul_Brennan@3com.com] Sent: mercredi, 21. janvier 1998 18:19 To: rvb@petrel.ch Subject: (usr-tc) Netserver 16/I doesn't authenticate (fwd) Dear Robert, we have opened a new case for in relation to this issue. The case number is 6677. An engineer will contact you as sonn as possible. Best regards Paul Brennan ---------------------- Forwarded by Paul Brennan/IE/3Com on 21/01/98 17:16 --------------------------- ---------- Forwarded message ---------- Date: Wed, 21 Jan 1998 16:31:27 +0100 From: Robert von Bismarck <rvb@petrel.ch> To: "'usr-tc@xmission.com'" <usr-tc@xmission.com> Subject: (usr-tc) Netserver 16/I doesn't authenticate I installed my Netserver 16/I + and have a problem. It doesn't even try to authenticate my dial-in attempts. Here is what it gives when I dial in with Hyperterminal or something like this=A0: OK atdtxxxxxxx CONNECT 115200 SPAN : Swiss Private Access Network NODE : COP-PM1 ACCESS : Restricted~#A!}!}!} }3}!}$}%=EA}#}%A#}%}%}&=B2j-G [--SNIP-SNIP--] I followed the CLI configuration wizard, and network is okay. It must be a simple configuration problem involving one or the other command, but I'm lost. Thank for any pointers, Robert von Bismarck Petrel Communication S.A. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.

I was curious if anyone can tell me where I can program a number for one modem to dial. Basically so that the modem can dial itself. I need to assign it the number and then have a script telnet in dial the number. ________________________________ / \ | Jennifer M Woodard | | Network Operations | | CyberGate Internet Technologies | | 1-954-429-8080 | \________________________________/

On Thu, 22 Jan 1998, Douglas Palmer wrote: > At 12:41 PM 1/22/98 -0500, you wrote: > > >And what you're hinting at is that...??? USR doesn't understand DOV > >calls? Anyone care to confirm or deny that? I can't see having ISDN in > >NYC without DOV capability. This feature to make it work, you must terminate the ISDN calls on the modem. Also you have to set the PRI to take a voice call and convert it to digital based on DNIS - this is done in the Inbound call routing menu of the pri. krish > > They'll answer them (DOV calls) alright, just no services once the line is > picked up. We may have to dump the USR rack for another system because > non-DOV ISDN is just too expensive in the NYC area, > > DCP > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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

Eagle will help you get PRI to work... They won't, however, make TCs do DOV even if they buy 3 more 5ESS switches.. > >Hello, > > > >We currently have 5 TC Hubs using channelized T1 service, and all is well. > >We just received an older chassis with a NMC card, Netserver card, and PRI > >card, and we'll be dedicating it to ISDN use. I wish to use it solely for > >ISDN, so there will be no modems installed in it. > > > >Having no experience with the PRI card or setting up the Netserver for > >ISDN, where is the best place to start? I've got an older (3.3) Netserver > >manual, and a ton of "hardware" manuals, but I've yet to find docs on the > >latest Netserver rev. Is there any type of "quick setup guide" for this? > >Can I do "data over voice" ISDN without the calls being routed to analog > >modems? > > > >Also, any suggestions on PRI provisioning are appreciated. Our telco is > >very accomodating and I can actually visit the CO and chat with the > >switch technicians once I know what to ask for... > > > >Thanks, > > > >Charles > > > >~~~~~~~~~ ~~~~~~~~~~~ > >Charles Sprickman Internet Channel > >INCH System Administration Team (212)243-5200 > >spork@inch.com access@inch.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. > > > > >

tc01.austin>version U.S. Robotics Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.6.28 Build date: Oct 2 1997 Build time: 14:42:06 Network Interface Card: Ethernet & Frame Relay Combination (26) ISDN Interface Card : MUNICH32 (4) Packet Bus Circuit : Standard tc01.austin>show mem Total physical RAM: 8192 kb Total physical FLASH: 2048 kb System memory 4195591 bytes - 4195440 used, 151 available Free blocks (block_size:count): 16:11 32:70 48:27 80:35 96:3 112:0 128:21 144:21 160:21 176:21 208:1 240:1 256:6 272:4 288:5 304:8 320:1 336:1 352:1 640:0 1168:0 1984:21 2064:2 4160:2 4208:3 8208:6 16400:5 20384:1 Real Available Memory: 246039 System nbufs 500 - 11 used, 489 available Is anyone else seeing a nasty memory leak in 3.6.28? I've found that every week or so our TC's need rebooting due to this problem and its getting rather annoying. --

> Total physical RAM: 8192 kb > Total physical FLASH: 2048 kb > > System memory 4195591 bytes - 4195440 used, 151 available > Free blocks (block_size:count): 16:11 32:70 48:27 80:35 96:3 112:0 128:21 > 144:21 160:21 176:21 208:1 240:1 256:6 272:4 288:5 304:8 320:1 336:1 352:1 > 640:0 1168:0 1984:21 2064:2 4160:2 4208:3 8208:6 16400:5 20384:1 > Real Available Memory: 246039 > System nbufs 500 - 11 used, 489 available > > Is anyone else seeing a nasty memory leak in 3.6.28? I've found that every > week or so our TC's need rebooting due to this problem and its getting > rather annoying. Although I've done it, I don't believe you are supposed to run 3.6.28 in 8mb of memory - you need to have 16mb for 3.6.28 - at least you did during the beta cycle of that code. However, to answer your question, I have seen a memory leak, but can't track anything down due to the fact that my netservers will go for months without a problem. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

At 08:54 AM 1/22/98 -0600, you wrote: >> Total physical RAM: 8192 kb >> Total physical FLASH: 2048 kb >> >> System memory 4195591 bytes - 4195440 used, 151 available >> Free blocks (block_size:count): 16:11 32:70 48:27 80:35 96:3 112:0 128:21 >> 144:21 160:21 176:21 208:1 240:1 256:6 272:4 288:5 304:8 320:1 336:1 352:1 >> 640:0 1168:0 1984:21 2064:2 4160:2 4208:3 8208:6 16400:5 20384:1 >> Real Available Memory: 246039 >> System nbufs 500 - 11 used, 489 available >> >> Is anyone else seeing a nasty memory leak in 3.6.28? I've found that every >> week or so our TC's need rebooting due to this problem and its getting >> rather annoying. > >Although I've done it, I don't believe you are supposed to run 3.6.28 in >8mb of memory - you need to have 16mb for 3.6.28 - at least you did during >the beta cycle of that code. > >However, to answer your question, I have seen a memory leak, but can't >track anything down due to the fact that my netservers will go for months >without a problem. > This is correct.. 3.6.28 is tested and approved for 16megs of RAM.. When there are out of memory issues it is quite possible for the NS to get very flakey.. You should up grade the RAM ASAP.. -m

This is a multi-part message in MIME format. --------------27B72CB6DFD1BEE34B766DB1 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The Double-Up offer actually doesn't end until Feb 15th of 98. Check out this page for info: http://www.3com.com/solutions/svprovider/double.html Jim Pergolizzi Autodesk Pete Ashdown wrote: > Jeff Binkley said once upon a time: > > >We are currently looking down the road at growth and the need to > >increase the number of ISDN Pri circuits beyond our current capacity. > >What I am struggling with is, how to do the upgrades ? The way I see > >it I have 3 choices with the TC racks. First we can take advantage > >of the double-up program and add two HyperDSP cards to our TC racks > >(Netserver & Quad digital cards). Secondly, we can buy more of the > >same TC racks or thirdly we can buy TC racks with Hyper Arc cards. > > Unfortunately, the Double-Up program is over. It ended December 15th. > However, I would recommend going HiPer in any future expansion. I have > resolved one major issue I had with the ARC, now I'm pushing another issue > through with the HDM cards. Other than that, I would recommend HiPer > because the performance is better, and the equipment seems to be structured > in a forward fashion rather than reverse. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. --------------27B72CB6DFD1BEE34B766DB1 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Pergolizzi, Jim Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Jim Pergolizzi n: Pergolizzi;Jim org: Autodesk, Inc. adr: 3900 Civic Center Dr.;;;San Rafael;CA;94903;USA email;internet: jim.pergolizzi@autodesk.com title: Network Engineer tel;work: 415-507-8897 tel;fax: 415-507-8395 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------27B72CB6DFD1BEE34B766DB1--

I thought 3.6.28 REQUIRED 16mb minimum to run. At 08:25 AM 1/22/98 -0600, you wrote: >tc01.austin>version >U.S. Robotics >Total Control (tm) NETServer Card V.34/ISDN with Frame Relay V3.6.28 > Build date: Oct 2 1997 > Build time: 14:42:06 > > Network Interface Card: Ethernet & Frame Relay Combination (26) > ISDN Interface Card : MUNICH32 (4) > Packet Bus Circuit : Standard > >tc01.austin>show mem >Total physical RAM: 8192 kb >Total physical FLASH: 2048 kb > >System memory 4195591 bytes - 4195440 used, 151 available >Free blocks (block_size:count): 16:11 32:70 48:27 80:35 96:3 112:0 128:21 >144:21 160:21 176:21 208:1 240:1 256:6 272:4 288:5 304:8 320:1 336:1 352:1 >640:0 1168:0 1984:21 2064:2 4160:2 4208:3 8208:6 16400:5 20384:1 >Real Available Memory: 246039 >System nbufs 500 - 11 used, 489 available > >Is anyone else seeing a nasty memory leak in 3.6.28? I've found that every >week or so our TC's need rebooting due to this problem and its getting >rather annoying. > >-- > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > >

Jeff Binkley said once upon a time: >We are currently looking down the road at growth and the need to >increase the number of ISDN Pri circuits beyond our current capacity. >What I am struggling with is, how to do the upgrades ? The way I see >it I have 3 choices with the TC racks. First we can take advantage >of the double-up program and add two HyperDSP cards to our TC racks >(Netserver & Quad digital cards). Secondly, we can buy more of the >same TC racks or thirdly we can buy TC racks with Hyper Arc cards. Unfortunately, the Double-Up program is over. It ended December 15th. However, I would recommend going HiPer in any future expansion. I have resolved one major issue I had with the ARC, now I'm pushing another issue through with the HDM cards. Other than that, I would recommend HiPer because the performance is better, and the equipment seems to be structured in a forward fashion rather than reverse.

And what you're hinting at is that...??? USR doesn't understand DOV calls? Anyone care to confirm or deny that? I can't see having ISDN in NYC without DOV capability. Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Thu, 22 Jan 1998, Garry Shtern wrote: > Date: Thu, 22 Jan 1998 02:45:16 -0500 (EST) > From: Garry Shtern <shterng@akula.com> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) PRI provisioning/setup > > > Eagle will help you get PRI to work... They won't, however, make TCs do > DOV even if they buy 3 more 5ESS switches.. > > > >Hello, > > > > > >We currently have 5 TC Hubs using channelized T1 service, and all is well. > > >We just received an older chassis with a NMC card, Netserver card, and PRI > > >card, and we'll be dedicating it to ISDN use. I wish to use it solely for > > >ISDN, so there will be no modems installed in it. > > > > > >Having no experience with the PRI card or setting up the Netserver for > > >ISDN, where is the best place to start? I've got an older (3.3) Netserver > > >manual, and a ton of "hardware" manuals, but I've yet to find docs on the > > >latest Netserver rev. Is there any type of "quick setup guide" for this? > > >Can I do "data over voice" ISDN without the calls being routed to analog > > >modems? > > > > > >Also, any suggestions on PRI provisioning are appreciated. Our telco is > > >very accomodating and I can actually visit the CO and chat with the > > >switch technicians once I know what to ask for... > > > > > >Thanks, > > > > > >Charles > > > > > >~~~~~~~~~ ~~~~~~~~~~~ > > >Charles Sprickman Internet Channel > > >INCH System Administration Team (212)243-5200 > > >spork@inch.com access@inch.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. > > > > > > > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

Brian said once upon a time: > >Can someone email me a filter they use on the Netserver for accounts that >are "mail" only. Basically I want the user to have access to ports 25 and >110 only of mail.shreve.net. I am going ot go ahead and do this now, I >just want to see what the rest of you are using so I don't leave anything >out. > >For example, would they need nameserver access? I don't think they would. >Are there any other "usefull" ports I should give them access too, this is >for a "mail only" account. They will be establishing a PPP connection. They do, to look up the IP address of your mail server. We also give them access to our web server, but nothing outside. I also give access to POP2, IMAP, and poppasswd ports. 198.60.22.2 is our primary name server, 198.60.22.22 is our mail/pop and secondary name server. Here's the filter I use on my netservers. I haven't translated it to ARC yet. add filter mail.in set filter mail.in 1 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 106 set filter mail.in 2 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 109 set filter mail.in 3 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 110 set filter mail.in 4 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 143 set filter mail.in 5 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 25 set filter mail.in 6 permit 0.0.0.0/0 198.60.22.2/32 udp dst eq 53 set filter mail.in 7 permit 0.0.0.0/0 198.60.22.22/32 udp dst eq 53 set filter mail.in 8 permit 0.0.0.0/0 198.60.22.2/32 tcp dst eq 53 set filter mail.in 9 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 53 set filter mail.in 10 permit 0.0.0.0/0 198.60.22.0/24 tcp dst eq 80 set filter mail.in 11 permit 0.0.0.0/0 198.60.22.0/24 tcp dst eq 8000 set filter mail.in 12 permit 0.0.0.0/0 198.60.22.0/24 tcp dst eq 443 set filter mail.in 13 deny 0.0.0.0/0 0.0.0.0/0 add filter mail.out set filter mail.out 1 permit 198.60.22.0/24 0.0.0.0/0 tcp estab set filter mail.out 2 permit 198.60.22.22/32 0.0.0.0/0 udp src eq 53 set filter mail.out 3 permit 198.60.22.2/32 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

At 12:41 PM 1/22/98 -0500, you wrote: >And what you're hinting at is that...??? USR doesn't understand DOV >calls? Anyone care to confirm or deny that? I can't see having ISDN in >NYC without DOV capability. They'll answer them (DOV calls) alright, just no services once the line is picked up. We may have to dump the USR rack for another system because non-DOV ISDN is just too expensive in the NYC area, DCP

At 12:50 PM 1/22/98 -0500, Douglas Palmer wrote: >At 12:41 PM 1/22/98 -0500, you wrote: > >They'll answer them (DOV calls) alright, just no services once the line is >picked up. We may have to dump the USR rack for another system because >non-DOV ISDN is just too expensive in the NYC area, We're seeing exactly the same problems. I've verified that the DNIS mapping in the PRI card is setup correctly. When I place a DOV call to the NETserver, various debug traces show that the call is interpreted as digital. So that part of the DOV feature is working correctly. However, PPP negotiation fails. A "sho isdn" command reveals that the channel in question is being interpreted as 64k instead of 56k. The strange thing is that on a DMS100 switch here in Atlanta, the NETserver seems to handle DOV properly (with quads, DSPs, or the internal Munich handling the call). Are you terminating ISDN calls on a Quad I-modem or on the NETserver directly? Do you see the same thing in "sho isdn"? One nice feature of Livingston PM-3 and Bay 5399/8000 gear is their ability to auto-sense DOV even without kludgy DNIS-based call routing. regards, -- Robert

At 12:50 PM 1/22/98 -0500, Douglas Palmer wrote: >At 12:41 PM 1/22/98 -0500, you wrote: > >>And what you're hinting at is that...??? USR doesn't understand DOV >>calls? Anyone care to confirm or deny that? I can't see having ISDN in >>NYC without DOV capability. > >They'll answer them (DOV calls) alright, just no services once the line is >picked up. We may have to dump the USR rack for another system because >non-DOV ISDN is just too expensive in the NYC area, > Here is a trick.. You DO NOT buy Total Controls to do ISDN in NYC. Use them for X2 support only. You guys have Ascends for ISDN and kflex. Just use USR for X2 and don't worry about ISDN on them. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

> >We are currently looking down the road at growth and the need to > >increase the number of ISDN Pri circuits beyond our current capacity. > >What I am struggling with is, how to do the upgrades ? The way I see > >it I have 3 choices with the TC racks. First we can take advantage > >of the double-up program and add two HyperDSP cards to our TC racks > >(Netserver & Quad digital cards). Secondly, we can buy more of the > >same TC racks or thirdly we can buy TC racks with Hyper Arc cards. > > Unfortunately, the Double-Up program is over. It ended December 15th. > However, I would recommend going HiPer in any future expansion. I have > resolved one major issue I had with the ARC, now I'm pushing another issue > through with the HDM cards. Other than that, I would recommend HiPer > because the performance is better, and the equipment seems to be structured > in a forward fashion rather than reverse. The Double-UP HDM program was extended to Feb 15th, as I am going to be purchasing a TON of them before the program ends... <grin> And, FWIW, you can get a HECK of a price per port by buying double up kits, and chassis and ARC cards all seperate. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

At 12:41 PM 1/22/98 -0500, Charles Sprickman wrote: >And what you're hinting at is that...??? USR doesn't understand DOV >calls? Anyone care to confirm or deny that? I can't see having ISDN in >NYC without DOV capability. > It's a well known fact. In several previous messages people have been discussing it. USR is promissing that to be released in the 1st Quarter of '98 along with OSPF support. But so far they can't even figure out the Quake lag bug :))) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

Can someone email me a filter they use on the Netserver for accounts that are "mail" only. Basically I want the user to have access to ports 25 and 110 only of mail.shreve.net. I am going ot go ahead and do this now, I just want to see what the rest of you are using so I don't leave anything out. For example, would they need nameserver access? I don't think they would. Are there any other "usefull" ports I should give them access too, this is for a "mail only" account. They will be establishing a PPP connection. 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 -----------------------------/

We are currently looking down the road at growth and the need to increase the number of ISDN Pri circuits beyond our current capacity. What I am struggling with is, how to do the upgrades ? The way I see it I have 3 choices with the TC racks. First we can take advantage of the double-up program and add two HyperDSP cards to our TC racks (Netserver & Quad digital cards). Secondly, we can buy more of the same TC racks or thirdly we can buy TC racks with Hyper Arc cards. From what I have seen 1 & 2 cost about the same and while I haven't seen pricing on #3, I suspect it costs more. The concerns with 1 & 2 are the long term supprot for Netservers and new features coming to that product set (i.e. IP voice etc..). Thus we would be investing in aging technology but that happens no matter when you buy technology. #2 has an advantage of reduced exposure to single points of failure and congestion (i.e. the netserver). I have real concerns about the Netserver's throughput on 96 connected channels. As long as space isn't a problem, #2 seems like a better choice. But the given length of life for the Netserver has me thinking the

We are currently looking down the road at growth and the need to increase the number of ISDN Pri circuits beyond our current capacity. What I am struggling with is, how to do the upgrades ? The way I see it I have 3 choices with the TC racks. First we can take advantage of the double-up program and add two HyperDSP cards to our TC racks (Netserver & Quad digital cards). Secondly, we can buy more of the same TC racks or thirdly we can buy TC racks with Hyper Arc cards. From what I have seen 1 & 2 cost about the same and while I haven't seen pricing on #3, I suspect it costs more. The concerns with 1 & 2 are the long term supprot for Netservers and new features coming to that product set (i.e. IP voice etc..). Thus we would be investing in aging technology but that happens no matter when you buy technology. #2 has an advantage of reduced exposure to single points of failure and congestion (i.e. the netserver). I have real concerns about the Netserver's throughput on 96 connected channels. As long as space isn't a problem, #2 seems like a better choice. But the given length of life for the Netserver has me thinking the new chassis is a better solution but I need to examine the costs closer. Any thoughts ? Jeff Binkley ASA Network Computing

Hi all.. Just looking for a sanity check here... I have a radius accounting program that I am trying to get to work. What I am trying to do is this: When a user disconnects, SNMP poll the modem they were connected to for connection information. I need to know, is how to arrive at the $slot from the $port int the radius accounting stop packet. $port = $p->attr('Client-Port-Id'); $port = $port + 3; $slot = $ports{$port}; $initial_TX_rate_snmp = "1.3.6.1.4.1.429.1.6.9.1.1.10.$slot" and then I had a table %ports = qw( 3 2001 4 2002 5 2003 6 2004 7 3001 8 3002 } but somehow I am managing to grab the data from the wrong modem. Any suggestions? Thanks, -Luke

On Thu, 22 Jan 1998, Garry Shtern wrote: > At 12:41 PM 1/22/98 -0500, Charles Sprickman wrote: > >And what you're hinting at is that...??? USR doesn't understand DOV > >calls? Anyone care to confirm or deny that? I can't see having ISDN in > >NYC without DOV capability. > > > > It's a well known fact. In several previous messages people have been > discussing it. USR is promissing that to be released in the 1st Quarter of > '98 along with OSPF support. But so far they can't even figure out the > Quake lag bug :))) Well I think someone on here troubeshoot it best by setting up a Linux box as the PPP server and still placing the call thru the USR TC. The totally eliminates the USR modems as being the problem, and totally places blame on the Netserver and/or PPP stack. I am still hoping they fix the Quake lag bug, it is no doubt the single biggest complaint (well that and ospf). 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 | | 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 -----------------------------/

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

Check to see how the modems are configured - the line interface options, the modem should be set to T1 TDM. krish On Thu, 22 Jan 1998, Brian wrote: > what are some possible problems with a cht1 line when you get ring no > answer. This is our first cht1 line, i have very little experience with > them, only PRI lines. > > Thanks, > > 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

Can someone from 3com/USR confirm/deny the following: DOVBS (Data over Voice bearer service) __does __does-not work on current Netserver code. I'm not talking about 64K DOV, just the nice simple 56K that all our customers expect. This is a lightly-loaded chassis with only a PRI, NMC, and Netserver card with the Munich daughtercard. No I-Modems, no need to take normal "analog" calls. Just want to route everything that comes in to the Netserver... I spoke to one USR tech today, but was not able to get a definitive answer. Thanks, C ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Thu, 22 Jan 1998, Douglas Palmer wrote: > Date: Thu, 22 Jan 1998 12:50:37 -0500 > From: Douglas Palmer <telos@gain-ny.com> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@lists.xmission.com > Subject: Re: (usr-tc) PRI provisioning/setup > > At 12:41 PM 1/22/98 -0500, you wrote: > > >And what you're hinting at is that...??? USR doesn't understand DOV > >calls? Anyone care to confirm or deny that? I can't see having ISDN in > >NYC without DOV capability. > > They'll answer them (DOV calls) alright, just no services once the line is > picked up. We may have to dump the USR rack for another system because > non-DOV ISDN is just too expensive in the NYC area, > > DCP > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

On Thu, 22 Jan 1998, Charles Sprickman wrote: > Can someone from 3com/USR confirm/deny the following: > > DOVBS (Data over Voice bearer service) __does __does-not work on current > Netserver code. I'm not talking about 64K DOV, just the nice simple 56K > that all our customers expect. > > This is a lightly-loaded chassis with only a PRI, NMC, and Netserver card > with the Munich daughtercard. No I-Modems, no need to take normal > "analog" calls. Just want to route everything that comes in to the > Netserver... I spoke to one USR tech today, but was not able to get a > definitive answer. > This is correct, you cannot terminate a DOV call directly on the NETServer. You need Quad-I modems to terminate ISDN Calls and you need to setup the PRI based on DNIS to the modems. This is how DOV works in our chassis. krish > Thanks, > > C > > ~~~~~~~~~ ~~~~~~~~~~~ > Charles Sprickman Internet Channel > INCH System Administration Team (212)243-5200 > spork@inch.com access@inch.com > > On Thu, 22 Jan 1998, Douglas Palmer wrote: > > > Date: Thu, 22 Jan 1998 12:50:37 -0500 > > From: Douglas Palmer <telos@gain-ny.com> > > Reply-To: usr-tc@lists.xmission.com > > To: usr-tc@lists.xmission.com > > Subject: Re: (usr-tc) PRI provisioning/setup > > > > At 12:41 PM 1/22/98 -0500, you wrote: > > > > >And what you're hinting at is that...??? USR doesn't understand DOV > > >calls? Anyone care to confirm or deny that? I can't see having ISDN in > > >NYC without DOV capability. > > > > They'll answer them (DOV calls) alright, just no services once the line is > > picked up. We may have to dump the USR rack for another system because > > non-DOV ISDN is just too expensive in the NYC area, > > > > DCP > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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. > \ 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

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 -----------------------------/

On Thu, 22 Jan 1998, System Administrator wrote: > On Thu, 22 Jan 1998, Brian wrote: > > > I am still hoping they fix the Quake lag bug, it is no doubt the single > > biggest complaint (well that and ospf). > > You and me both. At least there are multiple work-arounds w/ the lack of > OSPF. This latency issue (Quake being the biggest noticable symptom) is > just killing me here; and I haven't even started using my hdms yet. ISDN > udp latency DID get better under 3.6.69, when ISDN is terminated on the > Munich. I have had no luck terminating on Quad modems, varying widely > from 50 to 100ms just on standard icmp echo pings. Did anyone else notice this? I thought latency wasn't any better when terminating ISDN on the Munich than the Quads? Does MPIP work well with HDM's when your terminating on the Munich? what are the current drawbacks to terminating ISDN on the Munich, I remember there was a specific reason alot of us switched to terminating on the quads. > > I went over to a friends house the other day, who has his ISDN connected > to a provider using PM-3s. He's got an almost identical router as myself > on his end, and he sees 30ms icmps and gets __60-70__ Quakeworld pings on > the provider's server. Hell, I'm lucky to see a QW ping as low as 120. > The pm-3 in question was pretty heavily loaded too. 120 would be pushing it, on a tc hub its more like 200-1000 for 128 isdn. > > It's frustrating and it's a big money hole. If I could, I would switch to > Lucent/Livingston in a heartbeat. Definitely will NOT be purchasing > 3com/USR equipment in the future, unless this udp issue is resolved. IMO, > this _MUST_ be a software issue. Modern tcrs have *more* hardware > horsepower than the pm-3, yet can't keep up in this department. Again, > tres frustrating when you've spent tremendous amounts of money and time on > the best hardware and upstream bandwidth for your network, only to look > like the little mom-n'-pop ISP w/ a single tier-3 upstream and some crappy > old ascend router -- all because of your NAS boxen. > > Ok. Rant over. :) > > 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 -----------------------------/

On Thu, 22 Jan 1998, System Administrator wrote: > On Thu, 22 Jan 1998, Brian wrote: > > > I am still hoping they fix the Quake lag bug, it is no doubt the single > > biggest complaint (well that and ospf). > > You and me both. At least there are multiple work-arounds w/ the lack of > OSPF. This latency issue (Quake being the biggest noticable symptom) is > just killing me here; and I haven't even started using my hdms yet. ISDN > udp latency DID get better under 3.6.69, when ISDN is terminated on the > Munich. I have had no luck terminating on Quad modems, varying widely > from 50 to 100ms just on standard icmp echo pings. Did anyone else notice this? I thought latency wasn't any better when terminating ISDN on the Munich than the Quads? Does MPIP work well with HDM's when your terminating on the Munich? what are the current drawbacks to terminating ISDN on the Munich, I remember there was a specific reason alot of us switched to terminating on the quads. > > I went over to a friends house the other day, who has his ISDN connected > to a provider using PM-3s. He's got an almost identical router as myself > on his end, and he sees 30ms icmps and gets __60-70__ Quakeworld pings on > the provider's server. Hell, I'm lucky to see a QW ping as low as 120. > The pm-3 in question was pretty heavily loaded too. 120 would be pushing it, on a tc hub its more like 200-1000 for 128 isdn. > > It's frustrating and it's a big money hole. If I could, I would switch to > Lucent/Livingston in a heartbeat. Definitely will NOT be purchasing > 3com/USR equipment in the future, unless this udp issue is resolved. IMO, > this _MUST_ be a software issue. Modern tcrs have *more* hardware > horsepower than the pm-3, yet can't keep up in this department. Again, > tres frustrating when you've spent tremendous amounts of money and time on > the best hardware and upstream bandwidth for your network, only to look > like the little mom-n'-pop ISP w/ a single tier-3 upstream and some crappy > old ascend router -- all because of your NAS boxen. > > Ok. Rant over. :) > > 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 -----------------------------/

On Thu, 22 Jan 1998, Brian wrote: > I am still hoping they fix the Quake lag bug, it is no doubt the single > biggest complaint (well that and ospf). You and me both. At least there are multiple work-arounds w/ the lack of OSPF. This latency issue (Quake being the biggest noticable symptom) is just killing me here; and I haven't even started using my hdms yet. ISDN udp latency DID get better under 3.6.69, when ISDN is terminated on the Munich. I have had no luck terminating on Quad modems, varying widely from 50 to 100ms just on standard icmp echo pings. I went over to a friends house the other day, who has his ISDN connected to a provider using PM-3s. He's got an almost identical router as myself on his end, and he sees 30ms icmps and gets __60-70__ Quakeworld pings on the provider's server. Hell, I'm lucky to see a QW ping as low as 120. The pm-3 in question was pretty heavily loaded too. It's frustrating and it's a big money hole. If I could, I would switch to Lucent/Livingston in a heartbeat. Definitely will NOT be purchasing 3com/USR equipment in the future, unless this udp issue is resolved. IMO, this _MUST_ be a software issue. Modern tcrs have *more* hardware horsepower than the pm-3, yet can't keep up in this department. Again, tres frustrating when you've spent tremendous amounts of money and time on the best hardware and upstream bandwidth for your network, only to look like the little mom-n'-pop ISP w/ a single tier-3 upstream and some crappy old ascend router -- all because of your NAS boxen. Ok. Rant over. :) Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

On Thu, 22 Jan 1998, Brian wrote: > I am still hoping they fix the Quake lag bug, it is no doubt the single > biggest complaint (well that and ospf). You and me both. At least there are multiple work-arounds w/ the lack of OSPF. This latency issue (Quake being the biggest noticable symptom) is just killing me here; and I haven't even started using my hdms yet. ISDN udp latency DID get better under 3.6.69, when ISDN is terminated on the Munich. I have had no luck terminating on Quad modems, varying widely from 50 to 100ms just on standard icmp echo pings. I went over to a friends house the other day, who has his ISDN connected to a provider using PM-3s. He's got an almost identical router as myself on his end, and he sees 30ms icmps and gets __60-70__ Quakeworld pings on the provider's server. Hell, I'm lucky to see a QW ping as low as 120. The pm-3 in question was pretty heavily loaded too. It's frustrating and it's a big money hole. If I could, I would switch to Lucent/Livingston in a heartbeat. Definitely will NOT be purchasing 3com/USR equipment in the future, unless this udp issue is resolved. IMO, this _MUST_ be a software issue. Modern tcrs have *more* hardware horsepower than the pm-3, yet can't keep up in this department. Again, tres frustrating when you've spent tremendous amounts of money and time on the best hardware and upstream bandwidth for your network, only to look like the little mom-n'-pop ISP w/ a single tier-3 upstream and some crappy old ascend router -- all because of your NAS boxen. Ok. Rant over. :) Jesse Sipprell Senior Systems Engineer Evolution Communications, Inc. * Finger sysadmin@evcom.net for my PGP Public Key *

At 01:17 PM 1/22/98 -0500, Garry Shtern wrote: >At 12:41 PM 1/22/98 -0500, Charles Sprickman wrote: >>And what you're hinting at is that...??? USR doesn't understand DOV >>calls? Anyone care to confirm or deny that? I can't see having ISDN in >>NYC without DOV capability. >> > >It's a well known fact. In several previous messages people have been >discussing it. USR is promissing that to be released in the 1st Quarter of >'98 along with OSPF support. But so far they can't even figure out the >Quake lag bug :))) > OSPF in 68 days from now??.. On what? netserver, arc, or both?.. And will OSPF actually be out for the arc before MPIP??.. Any word on ARC/MPIP? Seeing how the netserver had MPIP long before OSPF (still doesn't), I would think the HiperARC would be the same case.. (or is it reverse engineered ;) 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

At 01:36 PM 1/22/98 -0500, Jeff Binkley wrote: > >We are currently looking down the road at growth and the need to >increase the number of ISDN Pri circuits beyond our current capacity. >What I am struggling with is, how to do the upgrades ? The way I see >it I have 3 choices with the TC racks. First we can take advantage >of the double-up program and add two HyperDSP cards to our TC racks >(Netserver & Quad digital cards). If we were not running any quake servers, then we would be very happy with the double-up! Unfortuately we are. If you host quake, your users cannot play on that chassis, even with ISDN.. And even if you don't host quake, your users might be playing over someone elses server.. I'm sure that quake lag + wan routing times is even more unplayable. So even if you are not a "quake house", your users might be playing quake in significant numbers.. (these quake clans spread like gangs..) >Secondly, we can buy more of the >same TC racks or thirdly we can buy TC racks with Hyper Arc cards. > >>From what I have seen 1 & 2 cost about the same We paid around a thousand dollars more for the double up than a 2059.. If we had purchased a 2059, we would not be having nearly as bad of a problem (but the lag is still there to a degree). >and while I haven't >seen pricing on #3, I suspect it costs more. Double-up: ~$13.5K Chassis: ~$3K HiperArc: ~$6K PowerSupplies: ?? what size.. My ballpark understanding makes me think that it is... >The concerns with >1 & 2 are the long term supprot for Netservers and new features >coming to that product set (i.e. IP voice etc..). Thus we would >be investing in aging technology but that happens no matter when >you buy technology. Like brian sez, what engineers to you think USR/3COM has working on the new stuff verses the old stuff? The Hiper line is definitely where the future is at... At least for "Quake Shops" and "ITSP's" anyway.. Therefore the delemma.. Do we live with half our modems crippled for Quake play. Or do we return the hdms and fall back to older technology.. #2 has an advantage of reduced exposure to >single points of failure and congestion (i.e. the netserver). I >have real concerns about the Netserver's throughput on 96 connected >channels. As long as space isn't a problem, #2 seems like a better >choice. > I am not a TC veteran. I would probably go with Pete's or David's opinion or someone who knows what they are talking about. But from my current experience, your decision revolves around your intensions regarding Quake Access and future stuff like IP Telephony.. 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

Quad I modem and the Quad modem are the same. Just make sure you have the latest Quad modem code on the Card. krish On Fri, 23 Jan 1998, Charles Sprickman wrote: > OK, > > So I'm still confused. I have a bunch of modem cards and have only heard > them reffered to as either quad digital or quad digital/analog. What is a > "quad I-modem"??? On the website > (http://www.3com.com/carrier/nsd/products/30412.html) there are > descriptions of available modems, and there is no mention of a "Quad > I-modem". Is this another name for a quad digital? If so, this will > actually work if I get a second number on the PRI trunk and use DNIS info > to route the DOV calls to a modem, right? > > arrrgh, > > Charles (who thinks it would be nice to have documentation on this sort > of thing) > > > > > This is correct, you cannot terminate a DOV call directly on the > > NETServer. You need Quad-I modems to terminate ISDN Calls and you need > > to setup the PRI based on DNIS to the modems. This is how DOV works in > > our chassis. > > > > krish > > > > > > > > > > > Thanks, > > > > > > C > > > > > > ~~~~~~~~~ ~~~~~~~~~~~ > > > Charles Sprickman Internet Channel > > > INCH System Administration Team (212)243-5200 > > > spork@inch.com access@inch.com > > > > > > On Thu, 22 Jan 1998, Douglas Palmer wrote: > > > > > > > Date: Thu, 22 Jan 1998 12:50:37 -0500 > > > > From: Douglas Palmer <telos@gain-ny.com> > > > > Reply-To: usr-tc@lists.xmission.com > > > > To: usr-tc@lists.xmission.com > > > > Subject: Re: (usr-tc) PRI provisioning/setup > > > > > > > > At 12:41 PM 1/22/98 -0500, you wrote: > > > > > > > > >And what you're hinting at is that...??? USR doesn't understand DOV > > > > >calls? Anyone care to confirm or deny that? I can't see having ISDN in > > > > >NYC without DOV capability. > > > > > > > > They'll answer them (DOV calls) alright, just no services once the line is > > > > picked up. We may have to dump the USR rack for another system because > > > > non-DOV ISDN is just too expensive in the NYC area, > > > > > > > > DCP > > > > > > > > - > > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > > with "unsubscribe usr-tc" in the body of the 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. > > > > > > > ----------------------------------------- > > \ 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. > > > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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

At 01:14 PM 1/22/98 -0600, Curt Shambeau wrote: >The Double-UP HDM program was extended to Feb 15th, as I am going to be >purchasing a TON of them before the program ends... <grin> Didn't know that there was an expiration to the program.. Sure wish my USR rep had my email address (and other customers) and gave us monthly updates on what's going on with their company.. Thanks for the tip. I need to accelerate my plans to purchase more HDM's if I can afford ARCs to go with them.. and can afford to loose some MPIP.. too many tradeoffs! >And, FWIW, you can get a HECK of a price per port by buying double up >kits, and chassis and ARC cards all seperate. > Could you elaborate on this a little.. My calculations seem to suggest otherwise.. Are HDM's cheaper "standalone" than in the "double-up" kit?.. 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

> 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. *ANY* netserver running 3.6.xx code will support up to 100 ports, period. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

what are some possible problems with a cht1 line when you get ring no answer. This is our first cht1 line, i have very little experience with them, only PRI lines. Thanks, 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 -----------------------------/

> >The Double-UP HDM program was extended to Feb 15th, as I am going to be > >purchasing a TON of them before the program ends... <grin> > > Didn't know that there was an expiration to the program.. Sure wish > my USR rep had my email address (and other customers) and gave us > monthly updates on what's going on with their company.. You probably just have to buy more... <grin> We purchase a couple million in Total Control and Access Builder each year, so I always have my rep's ear... <g> << HI JIM >> > >And, FWIW, you can get a HECK of a price per port by buying double up > >kits, and chassis and ARC cards all seperate. > > > > Could you elaborate on this a little.. My calculations seem to suggest > otherwise.. Are HDM's cheaper "standalone" than in the "double-up" kit?.. The double up kits are MUCH cheaper than buying standalone HDM's. That's why now is a great time to buy a full chassis of them. Buy 7 double up kits, a newer 130A chassis w/NMC card, a ARC card (or two - they load balance and act as redundancy). Now you have a 336 port box, and a great price per port. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

On Thursday, January 22, 1998 10:32 PM, Brian [SMTP:signal@shreve.net] wrote: > what are some possible problems with a cht1 line when you get ring no > answer. This is our first cht1 line, i have very little experience with > them, only PRI lines. Modem not on t1TDM Marshall Morgan http://www.netdoor.com 601.969.1434 Ext. #28 | Fax 601.969.3838 | 800.952.1570 Ext. #28

AMEN BROTHER!!! Why does it feel so good to know we are not the only ones who are suffering day in and day out! Thanks for the good company.. Allen At 08:55 PM 1/22/98 -0500, System Administrator wrote: >On Thu, 22 Jan 1998, Brian wrote: > >> I am still hoping they fix the Quake lag bug, it is no doubt the single >> biggest complaint (well that and ospf). > >You and me both. At least there are multiple work-arounds w/ the lack of >OSPF. This latency issue (Quake being the biggest noticable symptom) is >just killing me here; and I haven't even started using my hdms yet. ISDN >udp latency DID get better under 3.6.69, when ISDN is terminated on the >Munich. I have had no luck terminating on Quad modems, varying widely >from 50 to 100ms just on standard icmp echo pings. > >I went over to a friends house the other day, who has his ISDN connected >to a provider using PM-3s. He's got an almost identical router as myself >on his end, and he sees 30ms icmps and gets __60-70__ Quakeworld pings on >the provider's server. Hell, I'm lucky to see a QW ping as low as 120. >The pm-3 in question was pretty heavily loaded too. > >It's frustrating and it's a big money hole. If I could, I would switch to >Lucent/Livingston in a heartbeat. Definitely will NOT be purchasing >3com/USR equipment in the future, unless this udp issue is resolved. IMO, >this _MUST_ be a software issue. Modern tcrs have *more* hardware >horsepower than the pm-3, yet can't keep up in this department. Again, >tres frustrating when you've spent tremendous amounts of money and time on >the best hardware and upstream bandwidth for your network, only to look >like the little mom-n'-pop ISP w/ a single tier-3 upstream and some crappy >old ascend router -- all because of your NAS boxen. > >Ok. Rant over. :) > >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. > >

At 08:19 PM 1/22/98 -0600, 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. > Only two I believe with 12 quad cards (96 total modems). Probably four with no quads.. But why would you even want to know??.. We UDP lag *BAD* at 96 modems.. NO chance in hell of going even higher... Not without a new "miracle" code release anyway.. Allen

On Thu, 22 Jan 1998, Allen Marsalis wrote: > At 01:17 PM 1/22/98 -0500, Garry Shtern wrote: > >At 12:41 PM 1/22/98 -0500, Charles Sprickman wrote: > >>And what you're hinting at is that...??? USR doesn't understand DOV > >>calls? Anyone care to confirm or deny that? I can't see having ISDN in > >>NYC without DOV capability. > >> > > > >It's a well known fact. In several previous messages people have been > >discussing it. USR is promissing that to be released in the 1st Quarter of > >'98 along with OSPF support. But so far they can't even figure out the > >Quake lag bug :))) > > > > OSPF in 68 days from now??.. On what? netserver, arc, or both?.. And will > OSPF actually be out for the arc before MPIP??.. Any word on ARC/MPIP? > Seeing how the netserver had MPIP long before OSPF (still doesn't), I would > think the HiperARC would be the same case.. (or is it reverse engineered ;) Don't get excitied just yet :)) The way USR move they can't even get DOV to work on their PRI, not to mention Quake lag. Can anyone from 3com tell me exactly what is the status of this whole project. When should we expect solutions to the following problems: 1. DOV support in similar fashion PM3s 2. OSPF 3. Quake lag Please let me know what to expect and when to expect it. Thanks

On Thu, 22 Jan 1998, System Administrator wrote: > On Thu, 22 Jan 1998, Brian wrote: > > > I am still hoping they fix the Quake lag bug, it is no doubt the single > > biggest complaint (well that and ospf). > > You and me both. At least there are multiple work-arounds w/ the lack of > OSPF. This latency issue (Quake being the biggest noticable symptom) is > just killing me here; and I haven't even started using my hdms yet. ISDN > udp latency DID get better under 3.6.69, when ISDN is terminated on the > Munich. I have had no luck terminating on Quad modems, varying widely > from 50 to 100ms just on standard icmp echo pings. There is no normal solution to OSPF problem. OSPF has support for VLSM which RIP(v2) does not support and won't ever. Thanks to that routing tables from my PM3s are 10 times smaller than those of TCs.

On Thu, 22 Jan 1998, System Administrator wrote: > On Thu, 22 Jan 1998, Brian wrote: > > > I am still hoping they fix the Quake lag bug, it is no doubt the single > > biggest complaint (well that and ospf). > > You and me both. At least there are multiple work-arounds w/ the lack of > OSPF. This latency issue (Quake being the biggest noticable symptom) is > just killing me here; and I haven't even started using my hdms yet. ISDN > udp latency DID get better under 3.6.69, when ISDN is terminated on the > Munich. I have had no luck terminating on Quad modems, varying widely > from 50 to 100ms just on standard icmp echo pings. There is no normal solution to OSPF problem. OSPF has support for VLSM which RIP(v2) does not support and won't ever. Thanks to that routing tables from my PM3s are 10 times smaller than those of TCs.

Once upon a time Robert Sanders shaped the electrons to say... >One nice feature of Livingston PM-3 and Bay 5399/8000 gear is their ability >to auto-sense DOV even without kludgy DNIS-based call routing. Can you confirm that the Bay will do this? I've asked but no one ever replied before, and Bay doesn't indicate it on their lit. -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

--IMA.Boundary.356865588 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Luke, try subtracting 2 from your port instead of adding 3, or else reindex your table. The Client-Port-ID is the same value you'd see on the netserver (ie slot 2 channel 1 = S5). Steve Luke Gain <luke@erinet.com> AT Internet on 01/22/98 02:19:33 PM cc: --IMA.Boundary.356865588 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Description: cc:Mail note part Hi all.. Just looking for a sanity check here... I have a radius accounting program that I am trying to get to work. What I am trying to do is this: When a user disconnects, SNMP poll the modem they were connected to for connection information. I need to know, is how to arrive at the $slot from the $port int the radius accounting stop packet. $port = $p->attr('Client-Port-Id'); $port = $port + 3; $slot = $ports{$port}; $initial_TX_rate_snmp = "1.3.6.1.4.1.429.1.6.9.1.1.10.$slot" and then I had a table %ports = qw( 3 2001 4 2002 5 2003 6 2004 7 3001 8 3002 } but somehow I am managing to grab the data from the wrong modem. Any suggestions? Thanks, -Luke - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the 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.356865588 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 4C79EC51; Thu, 22 Jan 98 13:32:22 -0600 Received: from lists.xmission.com by usr.com (8.8.5/3.1.090690-US Robotics) id NAA27138; Thu, 22 Jan 1998 13:13:38 -0600 (CST) Received: from domo by lists.xmission.com with local (Exim 1.73 #4) id 0xvSF0-0006ov-00; Thu, 22 Jan 1998 12:24:42 -0700 Received: from erinet.com [207.0.229.27] by lists.xmission.com with esmtp (Exim 1.73 #4) id 0xvSEx-0006oC-00; Thu, 22 Jan 1998 12:24:39 -0700 Received: from eri.erinet.com (console.erinet.com [207.0.229.28]) by erinet.com (8.8.8/8.8.8.27) with ESMTP id OAA03325 for <usr-tc@lists.xmission.com>; Thu, 22 Jan 1998 14:26:22 -0500 (EST) Received: (from luke@localhost) by eri.erinet.com (8.8.5/8.8.0) id OAA26022 for usr-tc@lists.xmission.com; Thu, 22 Jan 1998 14:19:34 -0500 (EST) Message-Id: <199801221919.OAA26022@eri.erinet.com> X-Mailer: ELM [version 2.4 PL25] Content-Type: text Sender: owner-usr-tc@lists.xmission.com Precedence: bulk Reply-To: usr-tc@lists.xmission.com --IMA.Boundary.356865588--

On Thu, 22 Jan 1998, Marshall Morgan wrote: > On Thursday, January 22, 1998 10:32 PM, Brian [SMTP:signal@shreve.net] wrote: > > what are some possible problems with a cht1 line when you get ring no > > answer. This is our first cht1 line, i have very little experience with > > them, only PRI lines. > > Modem not on t1TDM > > Marshall Morgan I forgot to mention this is on a hdm, thus there is not t1TDM setting for a DTE source. Any other ideas? Brian > > > 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. > /-------------------------- 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 -----------------------------/

On Fri, 23 Jan 1998, Garry Shtern wrote: > > > On Thu, 22 Jan 1998, System Administrator wrote: > > > On Thu, 22 Jan 1998, Brian wrote: > > > > > I am still hoping they fix the Quake lag bug, it is no doubt the single > > > biggest complaint (well that and ospf). > > > > You and me both. At least there are multiple work-arounds w/ the lack of > > OSPF. This latency issue (Quake being the biggest noticable symptom) is > > just killing me here; and I haven't even started using my hdms yet. ISDN > > udp latency DID get better under 3.6.69, when ISDN is terminated on the > > Munich. I have had no luck terminating on Quad modems, varying widely > > from 50 to 100ms just on standard icmp echo pings. > > There is no normal solution to OSPF problem. OSPF has support for VLSM > which RIP(v2) does not support and won't ever. Thanks to that routing > tables from my PM3s are 10 times smaller than those of TCs. > RIPv2 does support VLSM, that is to say that subnet information is transmitted within the RIP updates....... > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Fri, 23 Jan 1998, Garry Shtern wrote: > > > On Thu, 22 Jan 1998, System Administrator wrote: > > > On Thu, 22 Jan 1998, Brian wrote: > > > > > I am still hoping they fix the Quake lag bug, it is no doubt the single > > > biggest complaint (well that and ospf). > > > > You and me both. At least there are multiple work-arounds w/ the lack of > > OSPF. This latency issue (Quake being the biggest noticable symptom) is > > just killing me here; and I haven't even started using my hdms yet. ISDN > > udp latency DID get better under 3.6.69, when ISDN is terminated on the > > Munich. I have had no luck terminating on Quad modems, varying widely > > from 50 to 100ms just on standard icmp echo pings. > > There is no normal solution to OSPF problem. OSPF has support for VLSM > which RIP(v2) does not support and won't ever. Thanks to that routing > tables from my PM3s are 10 times smaller than those of TCs. > RIPv2 does support VLSM, that is to say that subnet information is transmitted within the RIP updates....... > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

[Quoth Garry Shtern] ] ] ] On Thu, 22 Jan 1998, System Administrator wrote: ] ] > On Thu, 22 Jan 1998, Brian wrote: ] > ] > > I am still hoping they fix the Quake lag bug, it is no doubt the single ] > > biggest complaint (well that and ospf). ] > ] > You and me both. At least there are multiple work-arounds w/ the lack of ] > OSPF. This latency issue (Quake being the biggest noticable symptom) is ] > just killing me here; and I haven't even started using my hdms yet. ISDN ] > udp latency DID get better under 3.6.69, when ISDN is terminated on the ] > Munich. I have had no luck terminating on Quad modems, varying widely ] > from 50 to 100ms just on standard icmp echo pings. ] ] There is no normal solution to OSPF problem. OSPF has support for VLSM ] which RIP(v2) does not support and won't ever. Thanks to that routing ] tables from my PM3s are 10 times smaller than those of TCs. This was my biggest gripe with the TC, too. The Netserver's RIPv2 can be made to aggregate routes, and I've already shrunk my royting table by an order of magnitude. Mind you, I still want my OSPF, it's worlds nicer than RIP, and I like it on my Livingstons, Ascends, and Ciscos. Now this is smells of kludge, but it works, on rev 3.6.28 or better: . Put the net0 interfaces of your TC's in one subnet block. . Create subnet blocks for the dialup pools. . For the dialup pool, add an entry in the netmask table. "add netmask 169.207.64.64 255.255.255.192" . add an entry to the ippool table, even though this really isn't used "add ippool pool1 169.207.64.65 100" . turn on enh_routing . turn on ripv2 -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1

You can remove a couple of lines from this, DNS only uses tcp for zone transfers, so unless you have email only customers running their own DNS, you only need the udp entries for port 53. If you have mail only users running an smtp server, you may want to add: set filter mail.out x permit 0.0.0.0/0 0.0.0.0/0 tcp dst eq 25 (if you allow smtp from the outside world then it reduces the traffic on you smtp server slightly, as the mail can go directly to the customer) --- Russell Heilling -- NetDirect Internet Ltd Tel: +44 (0) 181 293 7000 Fax: +44 (0) 181 293 6000 email: russellh@netdirect.net.uk "In darkness we are all invisible" -----Original Message----- >Brian said once upon a time: >> >>Can someone email me a filter they use on the Netserver for accounts that >>are "mail" only. Basically I want the user to have access to ports 25 and >>110 only of mail.shreve.net. I am going ot go ahead and do this now, I >>just want to see what the rest of you are using so I don't leave anything >>out. >> >>For example, would they need nameserver access? I don't think they would. >>Are there any other "usefull" ports I should give them access too, this is >>for a "mail only" account. They will be establishing a PPP connection. > >They do, to look up the IP address of your mail server. We also give them >access to our web server, but nothing outside. I also give access to POP2, >IMAP, and poppasswd ports. > >198.60.22.2 is our primary name server, 198.60.22.22 is our mail/pop and >secondary name server. > >Here's the filter I use on my netservers. I haven't translated it to ARC >yet. > >add filter mail.in >set filter mail.in 1 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 106 >set filter mail.in 2 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 109 >set filter mail.in 3 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 110 >set filter mail.in 4 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 143 >set filter mail.in 5 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 25 >set filter mail.in 6 permit 0.0.0.0/0 198.60.22.2/32 udp dst eq 53 >set filter mail.in 7 permit 0.0.0.0/0 198.60.22.22/32 udp dst eq 53 >set filter mail.in 8 permit 0.0.0.0/0 198.60.22.2/32 tcp dst eq 53 >set filter mail.in 9 permit 0.0.0.0/0 198.60.22.22/32 tcp dst eq 53 >set filter mail.in 10 permit 0.0.0.0/0 198.60.22.0/24 tcp dst eq 80 >set filter mail.in 11 permit 0.0.0.0/0 198.60.22.0/24 tcp dst eq 8000 >set filter mail.in 12 permit 0.0.0.0/0 198.60.22.0/24 tcp dst eq 443 >set filter mail.in 13 deny 0.0.0.0/0 0.0.0.0/0 >add filter mail.out >set filter mail.out 1 permit 198.60.22.0/24 0.0.0.0/0 tcp estab >set filter mail.out 2 permit 198.60.22.22/32 0.0.0.0/0 udp src eq 53 >set filter mail.out 3 permit 198.60.22.2/32 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 > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

On Thu, 22 Jan 1998, Brett Hawn wrote: > Network Interface Card: Ethernet & Frame Relay Combination (26) > ISDN Interface Card : MUNICH32 (4) > Packet Bus Circuit : Standard > > tc01.austin>show mem > Total physical RAM: 8192 kb > Total physical FLASH: 2048 kb > > System memory 4195591 bytes - 4195440 used, 151 available > Free blocks (block_size:count): 16:11 32:70 48:27 80:35 96:3 112:0 128:21 > 144:21 160:21 176:21 208:1 240:1 256:6 272:4 288:5 304:8 320:1 336:1 352:1 > 640:0 1168:0 1984:21 2064:2 4160:2 4208:3 8208:6 16400:5 20384:1 > Real Available Memory: 246039 > System nbufs 500 - 11 used, 489 available > > Is anyone else seeing a nasty memory leak in 3.6.28? I've found that every > week or so our TC's need rebooting due to this problem and its getting > rather annoying. It may not be a memory leak. 3.6.28 says right on the Totalservice site that it requires 16MB of RAM on the Netserver. That's why I've held off upgrading my only 8MB Netserver to 3.6.28. Brian

Brian Elfert said once upon a time: > > > >On Thu, 22 Jan 1998, Michael Wronski wrote: > >> This is correct.. 3.6.28 is tested and approved for 16megs of RAM.. When >> there are >> out of memory issues it is quite possible for the NS to get very flakey.. >> You should >> up grade the RAM ASAP.. > >Does USR sell this RAM so I don't have to make 20 trips to the store to >find RAM that works? > >I've been meaning to upgrade my only 8MB Netserver. I believe it is 70 pin non-parity 70ns SIMMs. Pretty common stuff.

OK, So I'm still confused. I have a bunch of modem cards and have only heard them reffered to as either quad digital or quad digital/analog. What is a "quad I-modem"??? On the website (http://www.3com.com/carrier/nsd/products/30412.html) there are descriptions of available modems, and there is no mention of a "Quad I-modem". Is this another name for a quad digital? If so, this will actually work if I get a second number on the PRI trunk and use DNIS info to route the DOV calls to a modem, right? arrrgh, Charles (who thinks it would be nice to have documentation on this sort of thing) > > This is correct, you cannot terminate a DOV call directly on the > NETServer. You need Quad-I modems to terminate ISDN Calls and you need > to setup the PRI based on DNIS to the modems. This is how DOV works in > our chassis. > > krish > > > > > > Thanks, > > > > C > > > > ~~~~~~~~~ ~~~~~~~~~~~ > > Charles Sprickman Internet Channel > > INCH System Administration Team (212)243-5200 > > spork@inch.com access@inch.com > > > > On Thu, 22 Jan 1998, Douglas Palmer wrote: > > > > > Date: Thu, 22 Jan 1998 12:50:37 -0500 > > > From: Douglas Palmer <telos@gain-ny.com> > > > Reply-To: usr-tc@lists.xmission.com > > > To: usr-tc@lists.xmission.com > > > Subject: Re: (usr-tc) PRI provisioning/setup > > > > > > At 12:41 PM 1/22/98 -0500, you wrote: > > > > > > >And what you're hinting at is that...??? USR doesn't understand DOV > > > >calls? Anyone care to confirm or deny that? I can't see having ISDN in > > > >NYC without DOV capability. > > > > > > They'll answer them (DOV calls) alright, just no services once the line is > > > picked up. We may have to dump the USR rack for another system because > > > non-DOV ISDN is just too expensive in the NYC area, > > > > > > DCP > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the 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. > > > > ----------------------------------------- > \ 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. >

On Fri, 23 Jan 1998, Michael Mittelstadt wrote: > [Quoth Garry Shtern] > ] > ] > ] On Thu, 22 Jan 1998, System Administrator wrote: > ] > ] > On Thu, 22 Jan 1998, Brian wrote: > ] > > ] > > I am still hoping they fix the Quake lag bug, it is no doubt the single > ] > > biggest complaint (well that and ospf). > ] > > ] > You and me both. At least there are multiple work-arounds w/ the lack of > ] > OSPF. This latency issue (Quake being the biggest noticable symptom) is > ] > just killing me here; and I haven't even started using my hdms yet. ISDN > ] > udp latency DID get better under 3.6.69, when ISDN is terminated on the > ] > Munich. I have had no luck terminating on Quad modems, varying widely > ] > from 50 to 100ms just on standard icmp echo pings. > ] > ] There is no normal solution to OSPF problem. OSPF has support for VLSM > ] which RIP(v2) does not support and won't ever. Thanks to that routing > ] tables from my PM3s are 10 times smaller than those of TCs. > > This was my biggest gripe with the TC, too. The Netserver's RIPv2 can > be made to aggregate routes, and I've already shrunk my royting table > by an order of magnitude. Mind you, I still want my OSPF, it's worlds > nicer than RIP, and I like it on my Livingstons, Ascends, and Ciscos. > > Now this is smells of kludge, but it works, on rev 3.6.28 or better: > > . Put the net0 interfaces of your TC's in one subnet block. > . Create subnet blocks for the dialup pools. > . For the dialup pool, add an entry in the netmask table. > "add netmask 169.207.64.64 255.255.255.192" > . add an entry to the ippool table, even though this really > isn't used > "add ippool pool1 169.207.64.65 100" > . turn on enh_routing > . turn on ripv2 Michael, I would *love* to have this working. I remember when you initially conquered this task a few months back , you posted very clear instructions on how to do it, and I wasnt able to get it to work! 1. Can you use differnet pool sizes/subnet block sizes for the hubs? one of our hubs has 96 modems the others have 48, or does it all have to be the same? I will try this again and see if I can post the errors. Aggregating our routes saves MAJOR headaches, and cleans up a netstat -r greatly! I would think EVERYONE on this list was aggregating there routes, but from what I gather not many besides yourself have had success. Also its not like you can read about "aggregating your routes on the netserver" in any usr documentation. Brian > > -- > 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 | | 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 -----------------------------/

[Quoth Brian] ] On Fri, 23 Jan 1998, Michael Mittelstadt wrote: ] > . Put the net0 interfaces of your TC's in one subnet block. ] > . Create subnet blocks for the dialup pools. ] > . For the dialup pool, add an entry in the netmask table. ] > "add netmask 169.207.64.64 255.255.255.192" ] > . add an entry to the ippool table, even though this really ] > isn't used ] > "add ippool pool1 169.207.64.65 100" ] > . turn on enh_routing ] > . turn on ripv2 ] ] Michael, I would *love* to have this working. I remember when you ] initially conquered this task a few months back , you posted very clear ] instructions on how to do it, and I wasnt able to get it to work! ] ] 1. Can you use differnet pool sizes/subnet block sizes for the hubs? one ] of our hubs has 96 modems the others have 48, or does it all have to be ] the same? As long as each subnet is a valid CIDR network. For 48 modems, use a /26, for 96 modems, use a /25. I postulate that you'll need to do this anyway come that really wonderful day when OSPF for the TC is a reality. -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1

At 09:27 AM 1/23/98 -0600, Brian wrote: >On Fri, 23 Jan 1998, Garry Shtern wrote: >RIPv2 does support VLSM, that is to say that subnet information is >transmitted within the RIP updates....... > That is not VLSM, that is just subnet information. VLSM is what PM3s do, they take all the ip addresses in the pool and they divide them between subnets as such that you get only 5 routes to the entire pool instead of 48. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 09:27 AM 1/23/98 -0600, Brian wrote: >On Fri, 23 Jan 1998, Garry Shtern wrote: >RIPv2 does support VLSM, that is to say that subnet information is >transmitted within the RIP updates....... > That is not VLSM, that is just subnet information. VLSM is what PM3s do, they take all the ip addresses in the pool and they divide them between subnets as such that you get only 5 routes to the entire pool instead of 48. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Fri, 23 Jan 1998, Garry Shtern wrote: > At 09:27 AM 1/23/98 -0600, Brian wrote: > >On Fri, 23 Jan 1998, Garry Shtern wrote: > >RIPv2 does support VLSM, that is to say that subnet information is > >transmitted within the RIP updates....... > > > > That is not VLSM, that is just subnet information. VLSM is what PM3s do, > they take all the ip addresses in the pool and they divide them between > subnets as such that you get only 5 routes to the entire pool instead of 48. > > Garry Shtern shterng@akula.com > Chief Network Administrator http://www.akula.com > Akula Communications Corp. tel. (212) 292-8892 > Ok, maybe I am confused. RipV2 supports CIDR, and RIPv2 allows you to use variably lengthed subnet masks right? I mean whether its supports VLSM(tm) or not, you can use "variably lengthed subnet masks with ripv2" right? I have /32'2, /24's, /26's and /28's all going out over ripv2 from the netserver. I thought that since the netserver supports varibly lengthed subnet masks, that it by virtue is "VLSM", but it sounds like VLSM is a different beast? Can someone clear this up for me, I am a bit confused. 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 -----------------------------/

On Fri, 23 Jan 1998, Garry Shtern wrote: > At 09:27 AM 1/23/98 -0600, Brian wrote: > >On Fri, 23 Jan 1998, Garry Shtern wrote: > >RIPv2 does support VLSM, that is to say that subnet information is > >transmitted within the RIP updates....... > > > > That is not VLSM, that is just subnet information. VLSM is what PM3s do, > they take all the ip addresses in the pool and they divide them between > subnets as such that you get only 5 routes to the entire pool instead of 48. > > Garry Shtern shterng@akula.com > Chief Network Administrator http://www.akula.com > Akula Communications Corp. tel. (212) 292-8892 > Ok, maybe I am confused. RipV2 supports CIDR, and RIPv2 allows you to use variably lengthed subnet masks right? I mean whether its supports VLSM(tm) or not, you can use "variably lengthed subnet masks with ripv2" right? I have /32'2, /24's, /26's and /28's all going out over ripv2 from the netserver. I thought that since the netserver supports varibly lengthed subnet masks, that it by virtue is "VLSM", but it sounds like VLSM is a different beast? Can someone clear this up for me, I am a bit confused. 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 -----------------------------/

At 09:28 AM 1/23/98 -0600, Michael Mittelstadt wrote: >[Quoth Garry Shtern] >Now this is smells of kludge, but it works, on rev 3.6.28 or better: > >. Put the net0 interfaces of your TC's in one subnet block. >. Create subnet blocks for the dialup pools. >. For the dialup pool, add an entry in the netmask table. > "add netmask 169.207.64.64 255.255.255.192" >. add an entry to the ippool table, even though this really > isn't used > "add ippool pool1 169.207.64.65 100" >. turn on enh_routing >. turn on ripv2 > Don't you think it is pathetic that you have to go through all of this just to decrease the size of your routing table. Simple solution: 3com implements OSPF w/ VLSM. I have seriously hard time believing that Livingston is more advanced in a terminal server development than 3com. However, from all the bugs and unavailable features, it is starting to seem that way. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Fri, 23 Jan 1998, Michael Mittelstadt wrote: > [Quoth Brian] > ] On Fri, 23 Jan 1998, Michael Mittelstadt wrote: > ] > . Put the net0 interfaces of your TC's in one subnet block. > ] > . Create subnet blocks for the dialup pools. > ] > . For the dialup pool, add an entry in the netmask table. > ] > "add netmask 169.207.64.64 255.255.255.192" > ] > . add an entry to the ippool table, even though this really > ] > isn't used > ] > "add ippool pool1 169.207.64.65 100" > ] > . turn on enh_routing > ] > . turn on ripv2 > ] > ] Michael, I would *love* to have this working. I remember when you > ] initially conquered this task a few months back , you posted very clear > ] instructions on how to do it, and I wasnt able to get it to work! > ] > ] 1. Can you use differnet pool sizes/subnet block sizes for the hubs? one > ] of our hubs has 96 modems the others have 48, or does it all have to be > ] the same? > > As long as each subnet is a valid CIDR network. For 48 modems, use a > /26, for 96 modems, use a /25. I postulate that you'll need to do > this anyway come that really wonderful day when OSPF for the TC is a > reality. ok, understood, last question: After doing the above, can I just do a "reset net0" or do I have to reboot the beast? Brian > > -- > 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 | | 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 -----------------------------/

Was just wondering if any of you are going to ISPcon. Allen and myself will be going, and would be nice to meet with some of you and exchange some notes, put some faces with names etc. 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 -----------------------------/

> Was just wondering if any of you are going to ISPcon. Allen and myself > will be going, and would be nice to meet with some of you and exchange > some notes, put some faces with names etc. Myself and our V.P. of internet services (Michael Mittelstadt) will be coming to ISPCON. Hope to see you there... | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

On Fri, 23 Jan 1998, Garry Shtern wrote: > At 10:53 AM 1/23/98 -0600, Brian wrote: > >Ok, maybe I am confused. RipV2 supports CIDR, and RIPv2 allows you to > >use variably lengthed subnet masks right? I mean whether its supports > >VLSM(tm) or not, you can use "variably lengthed subnet masks with ripv2" > >right? I have /32'2, /24's, /26's and /28's all going out over ripv2 from > >the netserver. > > > >I thought that since the netserver supports varibly lengthed subnet masks, > >that it by virtue is "VLSM", but it sounds like VLSM is a different beast? > >Can someone clear this up for me, I am a bit confused. > > I was not talking about VLSM concept as is. I was talking about PM3s > ability to summarize all of your 96 IPs in about 5 routes. You mean "aggregate" your routes, which can be done on the hubs. Instead of sending out 48 /32 routes, the USR can be made to send out a single /26 route or whatever have you. 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 | | 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 -----------------------------/

On Fri, 23 Jan 1998, Garry Shtern wrote: > At 10:53 AM 1/23/98 -0600, Brian wrote: > >Ok, maybe I am confused. RipV2 supports CIDR, and RIPv2 allows you to > >use variably lengthed subnet masks right? I mean whether its supports > >VLSM(tm) or not, you can use "variably lengthed subnet masks with ripv2" > >right? I have /32'2, /24's, /26's and /28's all going out over ripv2 from > >the netserver. > > > >I thought that since the netserver supports varibly lengthed subnet masks, > >that it by virtue is "VLSM", but it sounds like VLSM is a different beast? > >Can someone clear this up for me, I am a bit confused. > > I was not talking about VLSM concept as is. I was talking about PM3s > ability to summarize all of your 96 IPs in about 5 routes. You mean "aggregate" your routes, which can be done on the hubs. Instead of sending out 48 /32 routes, the USR can be made to send out a single /26 route or whatever have you. 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 | | 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 -----------------------------/

On Thu, 22 Jan 1998, Michael Wronski wrote: > This is correct.. 3.6.28 is tested and approved for 16megs of RAM.. When > there are > out of memory issues it is quite possible for the NS to get very flakey.. > You should > up grade the RAM ASAP.. Does USR sell this RAM so I don't have to make 20 trips to the store to find RAM that works? I've been meaning to upgrade my only 8MB Netserver. Brian

> Does USR sell this RAM so I don't have to make 20 trips to the store to > find RAM that works? NON-Parity, NON-EDO 72-pin SIMM. It isn't that likely that a computer store (big chain at least) would have NON-EDO SIMMs lying around, but any smaller computer store would at least get them for you. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

At 10:53 AM 1/23/98 -0600, Brian wrote: >Ok, maybe I am confused. RipV2 supports CIDR, and RIPv2 allows you to >use variably lengthed subnet masks right? I mean whether its supports >VLSM(tm) or not, you can use "variably lengthed subnet masks with ripv2" >right? I have /32'2, /24's, /26's and /28's all going out over ripv2 from >the netserver. > >I thought that since the netserver supports varibly lengthed subnet masks, >that it by virtue is "VLSM", but it sounds like VLSM is a different beast? >Can someone clear this up for me, I am a bit confused. I was not talking about VLSM concept as is. I was talking about PM3s ability to summarize all of your 96 IPs in about 5 routes. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 10:53 AM 1/23/98 -0600, Brian wrote: >Ok, maybe I am confused. RipV2 supports CIDR, and RIPv2 allows you to >use variably lengthed subnet masks right? I mean whether its supports >VLSM(tm) or not, you can use "variably lengthed subnet masks with ripv2" >right? I have /32'2, /24's, /26's and /28's all going out over ripv2 from >the netserver. > >I thought that since the netserver supports varibly lengthed subnet masks, >that it by virtue is "VLSM", but it sounds like VLSM is a different beast? >Can someone clear this up for me, I am a bit confused. I was not talking about VLSM concept as is. I was talking about PM3s ability to summarize all of your 96 IPs in about 5 routes. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 11:09 AM 1/23/98 -0600, Brian wrote: > >You mean "aggregate" your routes, which can be done on the hubs. Instead >of sending out 48 /32 routes, the USR can be made to send out a single /26 >route or whatever have you. > >Brian But see with this scenario you have to bother with subneting on total controls. With pm3 you can do start the pool from whichever address you want and it will send the routing information according to your setup. You do not have to assign special netmasks which causes waste of IP space. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 11:09 AM 1/23/98 -0600, Brian wrote: > >You mean "aggregate" your routes, which can be done on the hubs. Instead >of sending out 48 /32 routes, the USR can be made to send out a single /26 >route or whatever have you. > >Brian But see with this scenario you have to bother with subneting on total controls. With pm3 you can do start the pool from whichever address you want and it will send the routing information according to your setup. You do not have to assign special netmasks which causes waste of IP space. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

Yeah those things take standard 72 pin memory, 60 or 70ns, non-parity, non-edo. Goes for about $15.00 At 10:32 AM 1/23/98 -0700, Pete Ashdown wrote: >Brian Elfert said once upon a time: >> >> >> >>On Thu, 22 Jan 1998, Michael Wronski wrote: >> >>> This is correct.. 3.6.28 is tested and approved for 16megs of RAM.. When >>> there are >>> out of memory issues it is quite possible for the NS to get very flakey.. >>> You should >>> up grade the RAM ASAP.. >> >>Does USR sell this RAM so I don't have to make 20 trips to the store to >>find RAM that works? >> >>I've been meaning to upgrade my only 8MB Netserver. > >I believe it is 70 pin non-parity 70ns SIMMs. Pretty common stuff. > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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

> > The Double-UP HDM program was extended to Feb 15th, as I am going to be > > purchasing a TON of them before the program ends... <grin> > > > > And, FWIW, you can get a HECK of a price per port by buying double up > > kits, and chassis and ARC cards all seperate. > > You can get a better price than buying a 2059 chassis for $11,500?? > > Isn't the double up around $13,000, a chassis around $4500, and the > HiperARC around $7000? OK, Let's compare. First, comparing a 2059 to a HiPer chassis is like comparing a 1984 Yugo to a 1997 BMW, but what the heck... It does the same job (getting you from place to place) but, well, you know... I've done these numbers with the thought that the minimum discount any ISP would settle for would be 30% off of list, and some may be getting 35%, so I put that number in there as well. If you get a better discount level than that, all the better. HiPer Chassis Loaded - 336 ports 7 Double-up kits: 134,428.00 1 HiPer ARC card: 9,995.00 1 HiPer Chassis 130A w/NMC 7,500.00 (2 power supplies 130A) Total: 151,923.00 Real ISP Price 30% disc 106,341.10 Real ISP Price 35% disc 98,749.95 Price per port 30% disc $ 316.50 Price per port 35% disc $ 293.89 2059 Chassis (1990 technology chassis with no upgrade path) Price: $ 11,500.00 No Discount Price per port: $ 239.58 1706/1866 Chassis (a real comparison) 34,400.00 (1 power supply 70A) Real ISP Price 30% discount 24,080.00 Real ISP Price 35% discount 22,360.00 Price per port 30% discount $ 501.66 Price per port 35% discount $ 465.83 BEST SOLUTION: Loaded HiPer Chassis w/336 ports & 2 load balancing and redundant HiPer ARC cards. 7 Double-up kits: 134,428.00 *2* HiPer ARC cards: 19,990.00 1 HiPer Chassis 130A w/NMC 7,500.00 (2 power supplies 130A) Total: 161,918.00 Real ISP Price 30% discount 113,342.60 Real ISP Price 35% discount 105,246.70 Price per port 30% discount $ 337.33 Price per port 35% discount $ 313.23 | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

yong_chen@nyed.uscourts.gov said once upon a time: > > > Urgent! We need a procedure to setting up a MPIP(Multilink PPP > Interspan Protocol) server in Unix. > > Do we need to install any software? If yes, what is it? As far as I have seen USR/3com has never released this server to the public. They talk about it in their documentation, but it isn't available. Can I ask why you need the UNIX server? The Netserver can act as one.

Urgent! We need a procedure to setting up a MPIP(Multilink PPP Interspan Protocol) server in Unix. Do we need to install any software? If yes, what is it? This is an emergency. Please help. Thanks much. Yong.

Once upon a time Lee Kuo shaped the electrons to say... >Does Lucent/Livingston still have a trade-in program? Not right now, I think that ended the end of December (or maybe November). However, as most people probably know, sales folks are there to make deals. Most vendors I've seen will make deals to varying degrees depending on the size of the potential sale. ie - talk to a sales person. -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

Randy Doran said once upon a time: > >Is anyone using the port-limit attribute in radius? I was hoping to use >it to restrict a user to 1 ISDN "B" channel. I set it to 1 for a test >user and the user was still able connect 2 "B" channels. Is this what >this attribute is used for? Or is there a better way to do this. Port-Limit works for us in the manner described. Are you on the latest code?

On Thu, 22 Jan 1998, System Administrator wrote: > It's frustrating and it's a big money hole. If I could, I would switch to > Lucent/Livingston in a heartbeat. Definitely will NOT be purchasing > 3com/USR equipment in the future, unless this udp issue is resolved. IMO, > this _MUST_ be a software issue. Modern tcrs have *more* hardware > horsepower than the pm-3, yet can't keep up in this department. Again, > tres frustrating when you've spent tremendous amounts of money and time on > the best hardware and upstream bandwidth for your network, only to look > like the little mom-n'-pop ISP w/ a single tier-3 upstream and some crappy > old ascend router -- all because of your NAS boxen. Does Lucent/Livingston still have a trade-in program?

On Fri, 23 Jan 1998, Garry Shtern wrote: > Don't you think it is pathetic that you have to go through all of this > just to decrease the size of your routing table. Simple solution: 3com > implements OSPF w/ VLSM. I have seriously hard time believing that > Livingston is more advanced in a terminal server development than 3com. > However, from all the bugs and unavailable features, it is starting to > seem that way. Livingston/3com comparison is not a good one. Try Lucent vs 3Com. Despite each having been bought out, I still think of them as their old names: Livingston & US Robotics. A better comparison, I would say.

On Thu, 22 Jan 1998, Tatai SV Krishnan wrote: > This is correct, you cannot terminate a DOV call directly on the > NETServer. You need Quad-I modems to terminate ISDN Calls and you need > to setup the PRI based on DNIS to the modems. This is how DOV works in > our chassis. Will the Quad-I modems do DOV calls over channelized T1 lines? I believe the PM3 does this currently.

Once upon a time Garry Shtern shaped the electrons to say... >to decrease the size of your routing table. Simple solution: 3com >implements OSPF w/ VLSM. I have seriously hard time believing that >Livingston is more advanced in a terminal server development than 3com. Simple point - USR licensed ComOS from Livingston *because* we were ahead. Remember, the NetServer is running (a now majorly mutated, granted) ComOS 3.1.4 base. If you aren't interested in what Lucent has coming, stop now...  As for more advanced... The HiPer is something our next product will more than sufficiently address, but I can't talk about that just yet... But keep in mind it will have all of the software features mentioned, and more, as well as being new HW. As for the TC vs PM-3. PM-3 is denser (modems per rack inch), has OSPF including NSSA, has BGP-4(!), outperforms the TC, can handle ISDN DATA, ISDN DOSBS, and modem calls on one number on a PRI - standard, can handle ISDN DOSBS and modem calls over CT1 - standard, the Lucent RADIUS server is on more platforms with more features, Lucent has ChoiceNet for centralized filter management, Stac is run in HW, and it runs much, much cooler. On T1 - PRI and CT1 dialin support at standard, with T1/fracT1/CT1 leased lines running PPP or Frame Relay standard. On E1 - PRI dialin support, with E1/fracE1/CE1 leased lines running PPP or Frame Relay. ComOS 3.8, now in beta, includes ethernet multihoming, OSPF over demand links, E1 R2 dialin (CE1), support for a new T1 WAN card (additional T1 port for leased line, E1 card pending), a new command subsytem, a new contextual help subsystem, expanded SNMP support including traps, Stac over leased lines, MP on modems, support for the new Pure Java PMConsole, BAP/BACP support, RADIUS auth based on CLID, route filters on demand lines, and a whole host of command improvements - like 'show route-to-dest' which will let you show just the route used for a specific destination, useful with large OSPF/BGP routing tables. The Java PMConsole introduces a host of features to the package itself. Using the command window you have command history and Perl-style regexp support, the ability to save and load configs as plaintext command files, a graqphical display of the system status (a representation of the back of the unit, with 'LEDs' for each modem, ether, etc), and several advanced debugging features. All in 100% Pure Java. ;-) And, of course, the new RADIUS ABM server is also 100% Pure Java. Undecided at this time is if 3.8 will support IPIP and IPSec - both now in beta. The PM-3 supports a HW daughter card for HW data encryption for use with IPSec. If not in 3.8 it will follow soon after. It really just depends on if we decide it is ready to ship when the rest of 3.8 is. Further out, now in lab testing, is NAT, NAPT (aka Proxy-NAT) and LSNAT (load sharing NAT). Note the current drafts on these are by a Lucent engineer. Also L2TP and NFAS are in active testing and are expected soon. Some might even make it in 3.8 depending on the testing program results. -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

yes. as soon as we put double up cards in our last 1706 chassis we are going to start with one hiper chassis with 1 hiper arc and add double up cards and another hiper arc card. even at $105,000-$113,000 this chassis would do approximately $67,200 ($200 x 336 ports!) and it would take up an amazing amount of space compared to 336 ports of 1706 chassis not to mention how much space courier v.everythings would take up with terminal servers and power supplies! matthew \At 01:56 PM 1/23/98 -0600, you wrote: >> > The Double-UP HDM program was extended to Feb 15th, as I am going to be >> > purchasing a TON of them before the program ends... <grin> >> > >> > And, FWIW, you can get a HECK of a price per port by buying double up >> > kits, and chassis and ARC cards all seperate. >> >> You can get a better price than buying a 2059 chassis for $11,500?? >> >> Isn't the double up around $13,000, a chassis around $4500, and the >> HiperARC around $7000? > >OK, Let's compare. First, comparing a 2059 to a HiPer chassis is like >comparing a 1984 Yugo to a 1997 BMW, but what the heck... It does the same >job (getting you from place to place) but, well, you know... > >I've done these numbers with the thought that the minimum discount any ISP >would settle for would be 30% off of list, and some may be getting 35%, so >I put that number in there as well. If you get a better discount level >than that, all the better. > >----------------------------------------------------------------------- >HiPer Chassis Loaded - 336 ports > >7 Double-up kits: 134,428.00 >1 HiPer ARC card: 9,995.00 >1 HiPer Chassis 130A w/NMC 7,500.00 (2 power supplies 130A) >-------------------------------------- >Total: 151,923.00 > >Real ISP Price 30% disc 106,341.10 >Real ISP Price 35% disc 98,749.95 > >Price per port 30% disc $ 316.50 >Price per port 35% disc $ 293.89 > >----------------------------------------------------------------------- > >2059 Chassis (1990 technology chassis with no upgrade path) >Price: $ 11,500.00 >No Discount >Price per port: $ 239.58 > >----------------------------------------------------------------------- > >1706/1866 Chassis (a real comparison) 34,400.00 (1 power supply 70A) > >Real ISP Price 30% discount 24,080.00 >Real ISP Price 35% discount 22,360.00 > >Price per port 30% discount $ 501.66 >Price per port 35% discount $ 465.83 > >----------------------------------------------------------------------- > >BEST SOLUTION: >Loaded HiPer Chassis w/336 ports & 2 load balancing and redundant >HiPer ARC cards. > >7 Double-up kits: 134,428.00 >*2* HiPer ARC cards: 19,990.00 >1 HiPer Chassis 130A w/NMC 7,500.00 (2 power supplies 130A) >------------------------------------------- >Total: 161,918.00 > >Real ISP Price 30% discount 113,342.60 >Real ISP Price 35% discount 105,246.70 > >Price per port 30% discount $ 337.33 >Price per port 35% discount $ 313.23 > >----------------------------------------------------------------------- > > > >-------------------------------------------------------------------------- >| 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. >

MPIP in the NetServer (for us at least) has been unreliable at best. This has wreaked havoc with our ISDN users, since every few days a clock goes out of sync or some other thing upsets the delicate balance apparently required for MPIP to happen. If anyone has any suggestions as to how to make it perform reliably, I'd certainly appreciate it. Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- yong_chen@nyed.uscourts.gov said once upon a time: > > > Urgent! We need a procedure to setting up a MPIP(Multilink PPP > Interspan Protocol) server in Unix. > > Do we need to install any software? If yes, what is it? As far as I have seen USR/3com has never released this server to the public. They talk about it in their documentation, but it isn't available. Can I ask why you need the UNIX server? The Netserver can act as one. - To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" with "unsubscribe usr-tc" in the body of the message. For information on digests or retrieving files and old messages send "help" to the same address. Do not use quotes in your message.

At 03:43 PM 1/23/98 -0600, Lee Kuo wrote: >On Fri, 23 Jan 1998, Garry Shtern wrote: >> Don't you think it is pathetic that you have to go through all of this >> just to decrease the size of your routing table. Simple solution: 3com >> implements OSPF w/ VLSM. I have seriously hard time believing that >> Livingston is more advanced in a terminal server development than 3com. >> However, from all the bugs and unavailable features, it is starting to >> seem that way. > >Livingston/3com comparison is not a good one. Try Lucent vs 3Com. >Despite each having been bought out, I still think of them as their old >names: Livingston & US Robotics. A better comparison, I would say. > The names changed but the people are still the same.. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

Is anyone using the port-limit attribute in radius? I was hoping to use it to restrict a user to 1 ISDN "B" channel. I set it to 1 for a test user and the user was still able connect 2 "B" channels. Is this what this attribute is used for? Or is there a better way to do this. Thanx, Randy Doran CyberGate Network Operations

At 02:06 AM 1/22/98 -0600, you wrote: >This feature to make it work, you must terminate the ISDN calls on the=20 >modem.=A0 Also you have to set the PRI to take a voice call and convert it= =20 >to digital based on DNIS - this is done in the Inbound call routing menu=20 >of the pri. As best as I can tell, that's how I've got it setup. When I dial in, I can= see the modem pick up via the TCM software and the client displays a connect (56000P) -- however, the DCD never goes high on the modem and the line eventually times out and hangs up. The quads have Sup. Rev 5.6.7, data pump ver 5.0.0, and sup. software date= of 7/24/97. This is the latest I could find on your site. Please show a sample of what the call routing status should display, as well as a sample PH=3D line from the PRI card setup. Do any of the programmable settings need to be modified in the Quads (via the TCM)? This is very frustrating.= :-\ DCP

On Fri, 23 Jan 1998, Garry Shtern wrote: > Yeah those things take standard 72 pin memory, 60 or 70ns, non-parity, > non-edo. Goes for about $15.00 I'll doubt you'll get a 16MB SIMM quite that cheap. I talked with Source Technology today, and they claimed to have a genuine USR 16MB SIMM for only $165! They also claimed that no other SIMM would work. I would pay maybe $75 for a USR SIMM, but $165 for one is outrageous. I'll take my chances on a generic one I guess. Brian

Well everyone, Allen finally went and got us a HyperARC card and another DoubleUp upgrade. Soooo, next week I got to get a HyperARC up and running with 24 Quads and 4 HDM's in it. Am I going to be hating life or is it not that bad? Eventually (like soon hopefully), this chassis will have 9 or 10 HDM's in it and "maybe" a second HyperARC. This will come as we move our 2059 bundles out to POPs and make our main site all Hyper technology. I have Pete's hyperarc configuration, and I am sure thats going to be the lifesaver (Thanks Pete! And if you have updated your config maybe repost it?) If anyone else has a hyperarc config if you could email it to me that would be great. I am going to download the HyperARC docs and get them bound ASAP. 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 -----------------------------/

Can I run 6 quads and 4 HDM's off: 1 70A 1 45A 2 70A 2 45A Any of those configurations? How much are the 130A supplies? /-------------------------- 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 -----------------------------/

On Fri, 23 Jan 1998, Peter D. Mayer wrote: > MPIP in the NetServer (for us at least) has been unreliable at best. This > has wreaked havoc with our ISDN users, since every few days a clock goes out > of sync or some other thing upsets the delicate balance apparently required > for MPIP to happen. If anyone has any suggestions as to how to make it > perform reliably, I'd certainly appreciate it. > I had clock problems at one time. The solution? Excercise your hardware contract and get the netserver replaced. You may have a buggy/flacky hardware rev, or just a fried clock. Also make sure you use a good timehost, i use time.nist.gov, which isnt that great, but it works. Brian > Peter D. Mayer > NetWalk Tech Support > dmayer@netwalk.com > > -----Original Message----- > From: Pete Ashdown <pashdown@xmission.com> > To: usr-tc@lists.xmission.com <usr-tc@lists.xmission.com> > Date: Friday, January 23, 1998 4:02 PM > Subject: Re: (usr-tc) MPIP Server > > > yong_chen@nyed.uscourts.gov said once upon a time: > > > > > > Urgent! We need a procedure to setting up a MPIP(Multilink PPP > > Interspan Protocol) server in Unix. > > > > Do we need to install any software? If yes, what is it? > > As far as I have seen USR/3com has never released this server to the > public. They talk about it in their documentation, but it isn't available. > > Can I ask why you need the UNIX server? The Netserver can act as one. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Fri, 23 Jan 1998, Randy Doran wrote: > Is anyone using the port-limit attribute in radius? I was hoping to use > it to restrict a user to 1 ISDN "B" channel. I set it to 1 for a test > user and the user was still able connect 2 "B" channels. Is this what > this attribute is used for? Or is there a better way to do this. Did you make sure radiusd re-read the users file? Make sure you restart radiusd and it re-reads the file. Yes Port-Limit works, we use it AND have tested it. Merit 2.4.23 > > Thanx, > Randy Doran > CyberGate Network Operations > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

On Fri, 23 Jan 1998, Randy Doran wrote: > Is anyone using the port-limit attribute in radius? I was hoping to use > it to restrict a user to 1 ISDN "B" channel. I set it to 1 for a test > user and the user was still able connect 2 "B" channels. Is this what > this attribute is used for? Or is there a better way to do this. Did you make sure radiusd re-read the users file? Make sure you restart radiusd and it re-reads the file. Yes Port-Limit works, we use it AND have tested it. Merit 2.4.23 > > Thanx, > Randy Doran > CyberGate Network Operations > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

Well that is what we paid for ours... As I mentioned, it is a standard SIMM. At 07:55 PM 1/23/98 -0600, Brian Elfert wrote: > > >On Fri, 23 Jan 1998, Garry Shtern wrote: > >> Yeah those things take standard 72 pin memory, 60 or 70ns, non-parity, >> non-edo. Goes for about $15.00 > >I'll doubt you'll get a 16MB SIMM quite that cheap. > >I talked with Source Technology today, and they claimed to have a genuine >USR 16MB SIMM for only $165! They also claimed that no other SIMM would >work. > >I would pay maybe $75 for a USR SIMM, but $165 for one is outrageous. >I'll take my chances on a generic one I guess. > >Brian > > Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Fri, 23 Jan 1998, Douglas C. Palmer wrote: > At 02:06 AM 1/22/98 -0600, you wrote: >=20 > >This feature to make it work, you must terminate the ISDN calls on the= =20 > >modem.=A0 Also you have to set the PRI to take a voice call and convert = it=20 > >to digital based on DNIS - this is done in the Inbound call routing menu= =20 > >of the pri. >=20 > As best as I can tell, that's how I've got it setup. When I dial in, I ca= n see > the modem pick up via the TCM software and the client displays a connect > (56000P) -- however, the DCD never goes high on the modem and the line > eventually times out and hangs up. >=20 > The quads have Sup. Rev 5.6.7, data pump ver 5.0.0, and sup. software dat= e of > 7/24/97. This is the latest I could find on your site. >=20 > Please show a sample of what the call routing status should display, as > well as > a sample PH=3D line from the PRI card setup. Do any of the programmable > settings > need to be modified in the Quads (via the TCM)? This is very frustrating.= :-\ >=20 > DCP To set this up you need to do setup on the PRI, the modem code that you=20 have is fine and has got nothing to do with this setup. What you need is= =20 to set up the pri - the inbound calling setup, where you have to setup=20 the PRI to send a DOV call to a modem/ group of modems. I do not have a=20 setup with me currently but will get you one monday. krish >=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 =09=09\=09T.S.V. Krishnan \ =09=09 \ Network System Engineer \ ( : - : ) =09=09 \ 3Com ............ \ =09=09----------------------------------------------/ tkrishna@bubba.ae.usr.com =20 ----------------------------/ http://interproc.ae.usr.com ----/ =09Any Sufficiently advanced bug is indistinguishable for a feature. =09=09=09=09=09=09- Rick Kulawiec

Brian, The setup is simple very simple. Go throught the Quick setup. That will setup almost everything for you. krish On Fri, 23 Jan 1998, Brian wrote: > Well everyone, Allen finally went and got us a HyperARC card and another > DoubleUp upgrade. > > Soooo, next week I got to get a HyperARC up and running with 24 Quads and > 4 HDM's in it. Am I going to be hating life or is it not that bad? > Eventually (like soon hopefully), this chassis will have 9 or 10 HDM's in > it and "maybe" a second HyperARC. This will come as we move our 2059 > bundles out to POPs and make our main site all Hyper technology. > > I have Pete's hyperarc configuration, and I am sure thats going to be the > lifesaver (Thanks Pete! And if you have updated your config maybe repost > it?) If anyone else has a hyperarc config if you could email it to me > that would be great. > > I am going to download the HyperARC docs and get them bound ASAP. > > 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

At 08:45 PM 1/23/98 -0600, Brian wrote: >Well everyone, Allen finally went and got us a HyperARC card and another >DoubleUp upgrade. Yeah we needed the modems and will do just about anything to rid ourselves of quake lag.. Couldn't wait any longer on USR's forthcoming arc upgrade for us netserver laggers.. (yes, there WILL be an upgrade someday.) I just hope I haven't replaced all my netservers before they do!.. Would be my luck after pushing so hard.. OK USR, I took my turn.. Spent $20K today.. Netserver #1 Quake lag - poof! Now it's your turn.. Can I please have a "Quake Lag Upgrade Kit" before I spend another $20K??... It would be most appreciated.. And most honorable.. 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

>At 08:45 PM 1/23/98 -0600, Brian wrote: >>Well everyone, Allen finally went and got us a HyperARC card and another >>DoubleUp upgrade. > >Yeah we needed the modems and will do just about anything to rid ourselves >of quake lag.. Couldn't wait any longer on USR's forthcoming arc upgrade >for us netserver laggers.. (yes, there WILL be an upgrade someday.) I just >hope I haven't replaced all my netservers before they do!.. Would be my luck >after pushing so hard.. > >OK USR, I took my turn.. Spent $20K today.. Netserver #1 Quake lag - poof! > >Now it's your turn.. Can I please have a "Quake Lag Upgrade Kit" before I >spend another $20K??... It would be most appreciated.. And most honorable.. > >allen > P.S. Anyone want to purchase a EPB Netserver (new)?? (I'm making my own damn quake lag upgrade kit!!) Any reasonable offer will be accepted and displayed here as "Current Quake Lag Upgrade Kit Pricing".. heh **Current Quake Lag Upgrade Kit Pricing** HiperARC $6,000 minus (reasonable offer) equals... around $5,000.. allen "Any problem can be fixed by beating it with your wallet!"

At 05:30 PM 1/23/98 -0500, you wrote: >Is anyone using the port-limit attribute in radius? I was hoping to use >it to restrict a user to 1 ISDN "B" channel. I set it to 1 for a test >user and the user was still able connect 2 "B" channels. Is this what >this attribute is used for? Or is there a better way to do this. Yes, and it works (on the NETserver, but not on the ARC -- they use the gratuitously different vendor-specific attribute number 0x9802). And it still won't keep them from *connecting*, of course, but it will keep them from bundling past the specified number of channels. -- Robert

On Sat, 24 Jan 1998, Cassandra M. Perkins wrote: > My USR-TC is only allowing one person to connect (on port S5). Any other > connection fails with the following error reported to syslogd: > > Jan 23 06:44:20 hpts01.loop.net S6 didn't get online! status=-1, > connect_fail=3 > 6, link_fail=31 This is a configurating issue, not a hardware problem. Check the configuration of the NETServer, and also make sure that dns is set up properlly. krish > > Does anyone know if this represents a hardware problem or a configuration > problem? > > > > ---------------------------------------------------------------------------- > | Cassandra M. Perkins | People usually get what's coming to | > | Network Operations | them... unless it's been mailed. | > | The Loop Internet Switch Co., LLC | -fortune | > ---------------------------------------------------------------------------- > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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

On Sat, 24 Jan 1998, Bob Roswell wrote: > Does this list cover USR Netserver/I? If not is there such a list? > > Bob Roswell > broswell@syssrc.com > There has been discussion the NETServer/I here before. I am not aware of a different list. You can certainly address the NETserver 8/16/I questions here. 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

well, ... I have been following the discussions on different TC racks and am a little confused. First off how does one tell what rack one has? Secound, what are the differences in performance for the different racks. As far as I can tell here there are 3 diffent racks. What are they? Also am I right that the hyperARC is the netserver replacement? I have an old non-enhanced Enterprise hub and 10 new ( I guess they are the old ones) enhanced packet bus TC's The old one has the built in fan tray and 70 amp power supply's, the new ( old? ) ones use dual 45 amp power supplies. The new ( old ) ones seem to be working fine. We only use them for X2 not ISDN as we are in USWEST land and PRi is not availible, well at least for a decent cost! How do the new chassis work better? I understand that they have larger power supplies and are ness. in order to use the Hyper cards, but what else gives. Is there actual processing power on the bus, is it faster? I thought the netserver was what powered the system in general anyways. I'm having a little buyers remorse right now. Did I make a mistake by buying the old bundles? They were quite a deal... I'm going to need more ports here soon, what do you all suggest to use? Should I go with the new chassis, 130 amp power supplies and hypercards and a hyperARC? or stay with what seems to be working? Well that's about 50 questions and probably 49 too many. Any help would be appreciated. Thanks, Terry Kennedy Olypen, Inc. NWiNET, Inc. terry@olypen.com www.olypen.com www.nwinet.com

My USR-TC is only allowing one person to connect (on port S5). Any other connection fails with the following error reported to syslogd: Jan 23 06:44:20 hpts01.loop.net S6 didn't get online! status=-1, connect_fail=3 6, link_fail=31 Does anyone know if this represents a hardware problem or a configuration problem? | Cassandra M. Perkins | People usually get what's coming to | | Network Operations | them... unless it's been mailed. | | The Loop Internet Switch Co., LLC | -fortune |

Does this list cover USR Netserver/I? If not is there such a list? Bob Roswell broswell@syssrc.com

Once upon a time Bob Roswell shaped the electrons to say... >Does this list cover USR Netserver/I? If not is there such a list? Yes. Since this is basically *the* 3Com/USR access server list it has always accepted issues on the NetServer products too. Fire away. -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

I setup the Netserver just like 2 other chassis I have. Only the one is reporting this problem. Where can I find a list of error codes reported that are reported to syslogd? | Cassandra M. Perkins | People usually get what's coming to | | Network Operations | them... unless it's been mailed. | | The Loop Internet Switch Co., LLC | -fortune | On Sat, 24 Jan 1998, Tatai SV Krishnan wrote: > On Sat, 24 Jan 1998, Cassandra M. Perkins wrote: > > > My USR-TC is only allowing one person to connect (on port S5). Any other > > connection fails with the following error reported to syslogd: > > > > Jan 23 06:44:20 hpts01.loop.net S6 didn't get online! status=-1, > > connect_fail=3 > > 6, link_fail=31 > > This is a configurating issue, not a hardware problem. Check the > configuration of the NETServer, and also make sure that dns is set up > properlly. > > krish > > > > > Does anyone know if this represents a hardware problem or a configuration > > problem? > > > > > > > > ---------------------------------------------------------------------------- > > | Cassandra M. Perkins | People usually get what's coming to | > > | Network Operations | them... unless it's been mailed. | > > | The Loop Internet Switch Co., LLC | -fortune | > > ---------------------------------------------------------------------------- > > > > > > - > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > with "unsubscribe usr-tc" in the body of the 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 > -------------------------------------------------------------------------/ > >

We have had good luck, better than with Sporsters Win modems with the X2 Global villages. We do, how ever, have all users set s10=200 in the modem init string and on the Carrier setting on the Quads in the TC. We buy them by the case for under $60 bucks each. Sure beats giving USR more money, we would still sell USR modems if they would only give us the same Tech support that Livingston does.

On Sat, 24 Jan 1998, Cassandra M. Perkins wrote: > I setup the Netserver just like 2 other chassis I have. Only the one is > reporting this problem. Where can I find a list of error codes reported > that are reported to syslogd? > > ---------------------------------------------------------------------------- > | Cassandra M. Perkins | People usually get what's coming to | > | Network Operations | them... unless it's been mailed. | > | The Loop Internet Switch Co., LLC | -fortune | > ---------------------------------------------------------------------------- > > On Sat, 24 Jan 1998, Tatai SV Krishnan wrote: > > > On Sat, 24 Jan 1998, Cassandra M. Perkins wrote: > > > > > My USR-TC is only allowing one person to connect (on port S5). Any other > > > connection fails with the following error reported to syslogd: > > > > > > Jan 23 06:44:20 hpts01.loop.net S6 didn't get online! status=-1, > > > connect_fail=3 > > > 6, link_fail=31 I do not have a list of all possible errors that you will see in the syslog. This error however tells me that the connection failed and it relates to a modem connection problem. Error 31 says that the remote side disconnected the call and is a modem link error. Check the configuration again, make sure that the modems are setup properlly, Use TCM to check the modem configuration. I would try to get one connection on s5 and then use a terminal program to dial and connect the second channel. Krish > > > > This is a configurating issue, not a hardware problem. Check the > > configuration of the NETServer, and also make sure that dns is set up > > properlly. > > > > krish > > > > > > > > Does anyone know if this represents a hardware problem or a configuration > > > problem? > > > > > > > > > > > > ---------------------------------------------------------------------------- > > > | Cassandra M. Perkins | People usually get what's coming to | > > > | Network Operations | them... unless it's been mailed. | > > > | The Loop Internet Switch Co., LLC | -fortune | > > > ---------------------------------------------------------------------------- > > > > > > > > > - > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > with "unsubscribe usr-tc" in the body of the 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 > > -------------------------------------------------------------------------/ > > > > > \ 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

I have a customer who has a Cardinal 340-56K Modem and having problems getting above 28.8 to my USR TC Rack. Does anyone have a good initialization string for this modem. Any help would be appreaciated. Thanks in advance. 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

I have a customer who has a Cardinal 340-56K Modem and having problems getting above 28.8 to my USR TC Rack. Does anyone have a good initialization string for this modem. Any help would be appreaciated. Thanks in advance. 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

I have a customer who has a Cardinal 340-56K Modem and having problems getting above 28.8 to my USR TC Rack. Does anyone have a good initialization string for this modem. Any help would be appreaciated. Thanks in advance. 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

I just read an article this morning about 56k modem problems accessing the TC hubs but I think they were only talking about USR Sportsters. It says that 3com posted updates on its website last autumn that contained bugs. The code is still there and 3com has no plans to remove it. The problem affects the modem fomerly called the Sportster. Two software updates dated Oct 3 and Oct 6, designed to speed modem performance , are instead blocking users from gaining access to service providers that support 3com's x2 56kbps format. Not everyone that downloads the upgrade experiences the problem. 3com has no plans to post a corrected update until Feb when it expects revamp the upgrade wizard to support the anticipated 56k standard. In the meantime, customers who downloaded the buggy upgrade can call 3com customer service at 847-982-5151 and ask for a file that will enable them to downgrade their modems to the previous version. 3com can be reached at 800-638-3266. This is news to me, wish I had heard about it in some other way than stumbling across it in a magazine article. At 06:40 PM 1/25/98 -0500, you wrote: > I have a customer who has a Cardinal 340-56K Modem and having problems >getting above 28.8 to my USR TC Rack. Does anyone have a good >initialization string for this modem. Any help would be appreaciated. >Thanks in advance. > >73's >John Campbell - KC4LWI >Owner - Roanoke Virginia Net >http://www.roava.net >mailto:sparky@roava.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. > > 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

Since s5 was working fine, I copied its configuration to all the other modems. It is now working fine. Thanks for your help. | Cassandra M. Perkins | People usually get what's coming to | | Network Operations | them... unless it's been mailed. | | The Loop Internet Switch Co., LLC | -fortune | On Sat, 24 Jan 1998, Tatai SV Krishnan wrote: > On Sat, 24 Jan 1998, Cassandra M. Perkins wrote: > > > I setup the Netserver just like 2 other chassis I have. Only the one is > > reporting this problem. Where can I find a list of error codes reported > > that are reported to syslogd? > > > > ---------------------------------------------------------------------------- > > | Cassandra M. Perkins | People usually get what's coming to | > > | Network Operations | them... unless it's been mailed. | > > | The Loop Internet Switch Co., LLC | -fortune | > > ---------------------------------------------------------------------------- > > > > On Sat, 24 Jan 1998, Tatai SV Krishnan wrote: > > > > > On Sat, 24 Jan 1998, Cassandra M. Perkins wrote: > > > > > > > My USR-TC is only allowing one person to connect (on port S5). Any other > > > > connection fails with the following error reported to syslogd: > > > > > > > > Jan 23 06:44:20 hpts01.loop.net S6 didn't get online! status=-1, > > > > connect_fail=3 > > > > 6, link_fail=31 > > > I do not have a list of all possible errors that you will see in the > syslog. This error however tells me that the connection failed and it > relates to a modem connection problem. Error 31 says that the remote > side disconnected the call and is a modem link error. > > Check the configuration again, make sure that the modems are setup > properlly, Use TCM to check the modem configuration. I would try to get > one connection on s5 and then use a terminal program to dial and connect > the second channel. > > Krish > > > > > > > > This is a configurating issue, not a hardware problem. Check the > > > configuration of the NETServer, and also make sure that dns is set up > > > properlly. > > > > > > krish > > > > > > > > > > > Does anyone know if this represents a hardware problem or a configuration > > > > problem? > > > > > > > > > > > > > > > > ---------------------------------------------------------------------------- > > > > | Cassandra M. Perkins | People usually get what's coming to | > > > > | Network Operations | them... unless it's been mailed. | > > > > | The Loop Internet Switch Co., LLC | -fortune | > > > > ---------------------------------------------------------------------------- > > > > > > > > > > > > - > > > > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > > > > with "unsubscribe usr-tc" in the body of the 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 > > > -------------------------------------------------------------------------/ > > > > > > > > > > ----------------------------------------- > \ 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 > -------------------------------------------------------------------------/ > >

Brian said once upon a time: >Soooo, next week I got to get a HyperARC up and running with 24 Quads and >4 HDM's in it. Am I going to be hating life or is it not that bad? >Eventually (like soon hopefully), this chassis will have 9 or 10 HDM's in >it and "maybe" a second HyperARC. This will come as we move our 2059 >bundles out to POPs and make our main site all Hyper technology. USR has fixed my routing problem with the latest engineering release. I only have one further complaint, and that is that I can't get the RADIUS "Idle-Timeout" to function at all. >I have Pete's hyperarc configuration, and I am sure thats going to be the >lifesaver (Thanks Pete! And if you have updated your config maybe repost >it?) If anyone else has a hyperarc config if you could email it to me >that would be great. Here's the latest. I should note that I used this after I did the setup Q&A when you first boot the HiPer. Something else that should be noted is that the TCM software download seems to be untrustworthy. Hook up a console, and reflash with ZModem at 115K. Here's the latest config, change the IP numbers to suit you: 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 add dns server 198.60.22.22 preference 2 set dns domain_name xmission.com set modem_group all message "Welcome to XMission Internet Access\r\n\r\n Type 'guest' for Guest Services\r\n\r\n 'account' for PPP\r\n 'account@slip' for SLIP\r\n 'account@shell' for Shell\r\n\r\n" set modem_group all prompt "slc10-tc login: " ## The following section must be done on a console, since you disable your ## network disable ip network ip set ip network ip routing_protocol ripv2 set ip network ip rip_policies_update no_ripv1_receive set ip network ip rip_policies_update no_send_compat enable ip network ip set authentication primary_secret SECRET set accounting primary_secret SECRET set system name XMission set system contact support@xmission.com set command prompt slc10-tc set syslog 198.60.22.5 facility log_local3 set ppp receive_authentication pap add ip pool slc10 initial 166.70.7.1 size 254 route aggregate enable ip security_option disallow_source_route_options set ntp primary_server 166.70.1.1 set ntp secondary_server 198.60.22.2

This probably isn't a fair question, but is it safe to try and get 11 HDMs and a single ARC running on a pair of 70 AMP power supplies? The reason I'd like to do this rather than the recommend 10 is that 253 channels would fit rather nicely into a class C subnet ip pool. If I decide to go to the 130 AMP power supplies, I'll probably use all 14 slots for HDMs to offset the costs of the power supplies. Also, if I put in a second ARC with 11 HDM's, will it overload the power supply?

> This probably isn't a fair question, but is it safe to try and get 11 HDMs > and a single ARC running on a pair of 70 AMP power supplies? The reason > I'd like to do this rather than the recommend 10 is that 253 channels would > fit rather nicely into a class C subnet ip pool. If I decide to go to the > 130 AMP power supplies, I'll probably use all 14 slots for HDMs to offset > the costs of the power supplies. > > Also, if I put in a second ARC with 11 HDM's, will it overload the power > supply? I've been told by USR (un-named sources) that the HDM cards actually take just a little more power than the old Quad cards. I hope this is true, as I've spent a lot on UPS's <grin> I've heard they built the 130A chassis in expectation of needing more power, but in the end, they may have been able to stick with the 70 chassis. I'm not sure what (if any) of this is true/false, but it's interesting. I think the best thing to do would be to plug a UPS in that has a load indicator on it. Plug in a fully loaded regular chassis, take a load reading. Plug in a HDM box, and see how close it is. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

For Sale: TOTAL CONTROL MP16/I MODEM ISDN & ANALOG MODEM POOL 16-LINES Has been in use for 5 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. Asking $4000. If interested or need more info, please email back privately. Thanks.

Charles Sprickman said once upon a time: > >Another good question is how would it run on *one* 70-Amp supply if one >were to fail. I think you'd lose your redundancy in that case... From what I've been able to understand, the power supplies don't work in tandem, but rather in a fail-over. Up until today, I've been running with 10 HDM's and an ARC on one 70 AMP power supply with no problems (power related problems that is).

After four weeks with the HDM's and the ARC's, I found several connectivity problems with these modem brands in particular=A0: - Hayes Ultra ISDN Terminal Adapter - Multitech ISDN Hopper Terminal Adapter - Multitech V34 Modems (ZDX model) - ZEUS 33600 Fax Modem (Rockwell based) (I bypassed the problem with disabling v.42bis but is this a good solution, is there a better one=A0?) Does anyone have any experience with these modems / TA's or with any other brands=A0? To be short, it would be useful, I think, to build a list of modems = that won't work with the USR's out of the box, and possible workarounds. I think I could post those workarounds on a web page somewhere as future reference for HDM using people. Second problem is, that the cards got stuck, and we had to do 'hardware reset' on the cards form the TCM. These problems only happened in one hub, so I'm beginning to suspect the chassis itself, even though it's = as new as the cards. Thanks for any replies, Robert von Bismarck Petrel Communication S.A.

After four weeks with the HDM's and the ARC's, I found several connectivity problems with these modem brands in particular=A0: - Hayes Ultra ISDN Terminal Adapter - Multitech ISDN Hopper Terminal Adapter - Multitech V34 Modems (ZDX model) - ZEUS 33600 Fax Modem (Rockwell based) (I bypassed the problem with disabling v.42bis but is this a good solution, is there a better one=A0?) Does anyone have any experience with these modems / TA's or with any other brands=A0? To be short, it would be useful, I think, to build a list of modems = that won't work with the USR's out of the box, and possible workarounds. I think I could post those workarounds on a web page somewhere as future reference for HDM using people. Second problem is, that the cards got stuck, and we had to do 'hardware reset' on the cards form the TCM. These problems only happened in one hub, so I'm beginning to suspect the chassis itself, even though it's = as new as the cards. Thanks for any replies, Robert von Bismarck Petrel Communication S.A.

It looks as if TCS 3.0.1 was released on the 23rd. It appears to mainly have new code for the NMC cards, new security/accounting servers, and new TCM managers.

Another good question is how would it run on *one* 70-Amp supply if one were to fail. I think you'd lose your redundancy in that case... Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Mon, 26 Jan 1998, Pete Ashdown wrote: > Date: Mon, 26 Jan 1998 13:16:48 -0700 (MST) > From: Pete Ashdown <pashdown@xmission.com> > Reply-To: usr-tc@lists.xmission.com > To: usr-tc@xmission.com > Subject: (usr-tc) 11 HDMs on 70 AMP? > > This probably isn't a fair question, but is it safe to try and get 11 HDMs > and a single ARC running on a pair of 70 AMP power supplies? The reason > I'd like to do this rather than the recommend 10 is that 253 channels would > fit rather nicely into a class C subnet ip pool. If I decide to go to the > 130 AMP power supplies, I'll probably use all 14 slots for HDMs to offset > the costs of the power supplies. > > Also, if I put in a second ARC with 11 HDM's, will it overload the power > supply? > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

I was starting to think that the entire web department of 3Com had been on a ship that sank at sea. However, it looks like some major updating has been going on with the Total Service site. I was able to look up and add to existing tickets! To whoever is in charge of this stuff, good work!

> I have a customer who has a Cardinal 340-56K Modem and having problems > getting above 28.8 to my USR TC Rack. Does anyone have a good > initialization string for this modem. Any help would be appreaciated. > Thanks in advance. > > 73's > John Campbell - KC4LWI > Owner - Roanoke Virginia Net > http://www.roava.net > mailto:sparky@roava.net > John (and the rest), I am using a 3440-56K Cardinal Internal modem from home and have not had a problem connecting to our TC. I get a 53,333K connect everytime without fail (I live right across the street from the CO) Thought I would share the info. Is this customer using a 340-56K, or a 3440-56K ? Ian Roy

US>For Sale: TOTAL CONTROL MP16/I MODEM ISDN & ANALOG MODEM POOL 16-LINES Has US>been in use for 5 months w/o any problems and currently in perfect working US>condition. X2 code has been enabled, was plugged into Portmaster-2E-30 and US>8 BRI lines to offer X2 connections. US>Asking $4000. If interested or need more info, please email back US>privately. Thanks. I'm somewhat interested but since you didn't leave an email address, I can't email you. Please respond to lhaggblad@mainline.ab.ca with your email address. Thanks! Len Haggblad MtnMail V3.05: A sure-fire formula for living to be one hundred - keep breathing!

We are disabling our POP and have the associated equipment for sale, to include a TC hub w/digital X2 quad modems, etc. Someone on one of these lists once posted a URL for ISP used equipment. Anyone remember the URL? Thanks, Jim

The modem is a 3440-56K... What Init Sting are you using on your cardinal. At 10:08 PM 1/26/98 -0700, you wrote: >> I have a customer who has a Cardinal 340-56K Modem and having problems >> getting above 28.8 to my USR TC Rack. Does anyone have a good >> initialization string for this modem. Any help would be appreaciated. >> Thanks in advance. >> >> 73's >> John Campbell - KC4LWI >> Owner - Roanoke Virginia Net >> http://www.roava.net >> mailto:sparky@roava.net >> > >John (and the rest), > > I am using a 3440-56K Cardinal Internal modem from home and have > not had a problem connecting to our TC. I get a 53,333K connect > everytime without fail (I live right across the street from the CO) > > Thought I would share the info. > > Is this customer using a 340-56K, or a 3440-56K ? > > Ian Roy > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

Trying to find a good place and price to but USR Total Control 2509 Chassis complete. I have heard you can get these chassis used for around 5-$6,000 Anyone have any ideas? Also anyone know how to set this chassis to knock off a user after a set time whether their idle or not? Jeff GlobalNet

I've noticed that USR changed the description for Netserver OS 3.6.28. The description now says that it'll work with 8MB or 16MB. It used to say just 8MB. Does anyone have 3.6.28 working 100% with just 8MB?. Brian

Other than the memory leaks I mentioned a while back it seems to be as reliable as any of their firmware is. On Tue, Jan 27, 1998 at 09:24:11AM -0600, Brian Elfert stated > I've noticed that USR changed the description for Netserver OS 3.6.28. > > The description now says that it'll work with 8MB or 16MB. It used to say > just 8MB. > > Does anyone have 3.6.28 working 100% with just 8MB?. > > 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. --

Robert, Bob Bockman MCI, I have had experience with the rockwell modem set in v.42 and have also found that turning off v.42 works and also forcing mnp4/5 works as well. I have worked with USR and in modem code version 5.6.7 / 5.5.7 this problem is addressed. I have not had this problem with the Rockwell chipset since this release. This release came out in tcs2.5.1. Unfortunately, I have not been exposed to the other 3 modems/isus in this emial. Hopes this helps bob

> The modem is a 3440-56K... What Init Sting are you using on your cardinal. > John, I am using it under windows95 and have not done anything special (ie. no additional init strings). I can look it up tonight and let you know. PS: I have been successfully using it with Win31 and also Arachne (the DOS graphical web browser). Thanks Ian

Have your RADIUS send attribute 27 (Session-Timeout) with a value in seconds (i.e. 43200 seconds = 12 hours). I don't know if USR's RADIUS supports sending this or not, but it is a standard RFC attribute. Peter D. Mayer NetWalk Tech Support dmayer@netwalk.com -----Original Message----- >Also anyone know how to set this chassis to knock off a user after a set >time >whether their idle or not? > >Jeff >GlobalNet > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

I have a 4MB NMC card with 4.3.4 on it, that has a write password that somehow is not in line with all my other ones. I popped the card, set the CLR jumper, popped it back in, red lights. Waited a few seconds, popped it back out and reset to NORM, and it's still got the old configuration in it. I have also flipped switch 5 to boot from factory defaults, and that seems to make no difference either. Any tip appreciated. I'm going to call USR support, but since it will be 6 hours on hold, I don't feel like doing that until tomorrow.

On Tue, 27 Jan 1998, cox wrote: > We are disabling our POP and have the associated equipment for sale, to > include a TC hub w/digital X2 quad modems, etc. > > Someone on one of these lists once posted a URL for ISP used equipment. > Anyone remember the URL? isp-services@ispc.org > > Thanks, > > Jim > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

Does anyone know what this means when it flies accross your HyperARC console: At 19:12:54, Facility "Call Initiation Process", Leve "CRITICAL":: CIP Set the type to dial_out in the user record Just wondering 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 -----------------------------/

Ok, Our HyperARC has been up for about 2 hours now, with very little problems, much thanks to Pete Ashdown for letting me take a look at his setup. To get it "up and running" was not that difficult. So far I have not experienced any problems. The new OS definitly seems lightyears ahead of ComOS, and *initial* quake lag tests seem MUCH better. I can get 100-120 ping times on 128k isdn, which isn't extremley good, but thats just *one* test, I will let you all know what other ISDN/dialup users report as quake ping times. I like the two 100bt ports. Can someone tell me if they are capible of load balancing or if they are strictly for redundancy? I am running the latest release code on my HyperARC and HyperDSP's. I heard there was a way to access the HyperDSP's from within the HyperARC, anyone care to comment on this? Forgive me if the question is a simple one, but I have only been running this HyperARC for a few hours and I am very anxious to learn as much as I can. My initial feelings are positive, and I know we made the right move. I am *sure* there are bugs/problems/compatibilty issues and I will fight those in the upcoming months, but the netserver had those too, and I am confident they get fixed fast :). We also aren't pushing the HyperARC too much, We only have about 125 calls going to it right now. does anyone have any comments on HyperARC vs. REALLY loaded HyperARC? in other words do you find its a "must do" to put dual HyperARC's in a hub with x number of HyperDSP modems? Our plan is to slam chassis with as many HyperDSP's as they will take, and put dual HyperARC's in them, but what I am wondering, is it really necessary or is performance hit marginal? 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 -----------------------------/

Thus spake Jaye Mathisen >Any tip appreciated. I'm going to call USR support, but since it will be >6 hours on hold, I don't feel like doing that until tomorrow. Actually, to 3Com's credit, the hold times for tech support *have* gotten shorter (or is it just my imagination?), though the expertise level at the other end is still a bit lacking. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

On Tue, 27 Jan 1998, Brian wrote: > Ok, Our HyperARC has been up for about 2 hours now, with very little > problems, much thanks to Pete Ashdown for letting me take a look at his > setup. > > To get it "up and running" was not that difficult. So far I have not > experienced any problems. The new OS definitly seems lightyears ahead of > ComOS, and *initial* quake lag tests seem MUCH better. I can get 100-120 > ping times on 128k isdn, which isn't extremley good, but thats just *one* > test, I will let you all know what other ISDN/dialup users report as quake > ping times. > > I like the two 100bt ports. Can someone tell me if they are capible of > load balancing or if they are strictly for redundancy? It is not for load balancing, some use one port for user data and the other for management. > > I am running the latest release code on my HyperARC and HyperDSP's. > > I heard there was a way to access the HyperDSP's from within the HyperARC, > anyone care to comment on this? Forgive me if the question is a simple > one, but I have only been running this HyperARC for a few hours and I am > very anxious to learn as much as I can. > You can set the modems and send parameters to the modem but if you are talking about some feture like logging on the the console port - that feature will be present in the next major release of code. > My initial feelings are positive, and I know we made the right move. I am > *sure* there are bugs/problems/compatibilty issues and I will fight those > in the upcoming months, but the netserver had those too, and I am > confident they get fixed fast :). > > We also aren't pushing the HyperARC too much, We only have about 125 calls > going to it right now. > > does anyone have any comments on HyperARC vs. REALLY loaded HyperARC? in > other words do you find its a "must do" to put dual HyperARC's in a hub > with x number of HyperDSP modems? > Each HiPer ARC with the current code supports 7 HDM - you can have two HiPer ARC in one chassis, but make sure that you disable chassis awarness on the HiPer ARCS. "disable nmc chas " Also you must configure each HiPer ARC to be owner of only 7 HDM. > Our plan is to slam chassis with as many HyperDSP's as they will take, and > put dual HyperARC's in them, but what I am wondering, is it really > necessary or is performance hit marginal? > > 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

I had to call twice today and didn't even wait 30 seconds before I got a tech. James >Thus spake Jaye Mathisen >>Any tip appreciated. I'm going to call USR support, but since it will be >>6 hours on hold, I don't feel like doing that until tomorrow. > >Actually, to 3Com's credit, the hold times for tech support *have* >gotten shorter (or is it just my imagination?), though the expertise >level at the other end is still a bit lacking. >-- >Jeff McAdams Email: jeffm@iglou.com

On Wed, 28 Jan 1998, Pete Ashdown wrote: > Tatai SV Krishnan said once upon a time: > > >Each HiPer ARC with the current code supports 7 HDM - you can have two > >HiPer ARC in one chassis, but make sure that you disable chassis awarness > >on the HiPer ARCS. "disable nmc chas " > >Also you must configure each HiPer ARC to be owner of only 7 HDM. > > Why am I running 10? Or does this apply to chassis with two ARCs? The HiPer ARC is capable of having 10 HDM in a chassis. The current release of code - we suggest users to upto 7 modems. Also you can have 2 HiPer ARC in one chassis. 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

I have a question regarding how the HyperARC sends out its routing updates: 1. When a user dials into a HyperARC chassis, if the user uses dynamic IP assignment, and is assigned an address out of the HyperARC's ippool, then the HyperARC sends out its aggregated route. So if I have an ippool such as: IP ADDRESS POOLS Name Address Size State Route Status pool1 208.214.044.001 /H 254 PUBLIC NO_AGGREGATE ACTIVE Then as soon as the first user dials in a route is sent out to all servers a la: 208.214.44.0 usr1ts1.shreve. 255.255.255.0 UG 1500 eth0 (usr1ts1 is my HyperARC). This is no problem. However, if a user is assigned a Framed-Route by RADIUS, which is not in an ippool on the HyperARC, then it would appear that the HyperARC does not send out a RIP update, because my UNIX servers, running gated(ripv2), don't see anything. If a user with: Framed-Route "208.214.45.0/28 208.214.45.1 1" dials in, I expect to see: 208.214.45.0 208.214.45.1 255.255.255.240 UG 1500 eth0 Here is how my dynamic routing is declared on the HyperARC: set ip network ip routing_protocol ripv2 set ip network ip rip_policies_update no_ripv1_receive set ip network ip rip_policies_update no_send_compat So basically we want it to send/receive ripv2, which its doing because I am seeing the route for the ippool, just not Framed-Route users routes. Has anyone else seen this? Basically I want the HyperARC to behave EXACTLY like a Netserver would with ripv2/send/receive turned on. On the netserver, ANY routing requests out of the pool will send an aggregated route out (if enh_routing is on), and thats being done. Also any Framed-Route users will cause a route to go out as a ripv2 update. Now......my network above: "ip", is not the 208.214.45.0 network, do I have to declare a network name for ALL network blocks I have and send out those statments above to turn on ripv2 updates for that network? This would make sense to me. disable ip network ip set ip network ip routing_protocol ripv2 set ip network ip rip_policies_update no_ripv1_receive set ip network ip rip_policies_update no_send_compat enable ip network ip HiPer>> show network ip SHOW IP NETWORK ip SETTINGS: Interface: eth:1 Network Address: 208.206.76.35/C Frame Type: ETHERNET_II Status: ENABLED Reconfigure Needed: FALSE Mask: 255.255.255.0 Station: 208.206.76.35 Broadcast Algorithm: 1 Max Reassembly Size: 3464 IP Routing Protocol: RIPV2 IP RIP Routing Policies: SEND_ROUTES SPLIT_HORIZON FLASH_UPDATE RIPV2_RECEIVE The above will send out ripv2 routes for 208.214.44.0, why not a framed-route from 208.214.45.0? /-------------------------- 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 -----------------------------/

> The modem is a 3440-56K... What Init Sting are you using on your cardinal. > John, Last night I tried the following init string - ATZ and everything worked fine. I was connecting at 53,333 bps Thanks Ian

In the never ending drama of setting up our first tc hub... Well, we finaly can connect to the thing. ICG has set the PRI to send 7 digits on the D channel instead of 0, and now there are no more fast busies. However, we can only connect at 9600 baud. What do I need to check to get this working right? James

Putting together our first fully loaded HiPer Chassis. However, the 130A power supplies are on back order. Loaded up a 70A chassis with 14 HDM cards and 2 HiPer ARC cards. No problem for even 1 70A power supply as of yet... I'll try and do some more real world load readings ASAP. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

Curt Shambeau said once upon a time: >Putting together our first fully loaded HiPer Chassis. However, the 130A >power supplies are on back order. Loaded up a 70A chassis with 14 HDM >cards and 2 HiPer ARC cards. No problem for even 1 70A power supply as >of yet... Verrrry interesting. I wonder what the 130A power supplys are all about then? >I'll try and do some more real world load readings ASAP. That would be very cool.

Robert von Bismarck said once upon a time: >- ISDN does connect with lots of difficulties, I had to really >fiddle with my trusty Elite2864i (Zyxel) and it connects OK now, but I >can't expect this from my clients. I snooped one the connections using >the "monitor PPP" feature from the HiPerARC, and got some weird tracing >messages (I'll send anyone interested this logfile). Here are the settings I use with my HDM's ISDN section. Please note two things about HDM ISDN. First, "v120 enabled" does not "stick" as a setting. I think it is being set properly, but being read in the wrong fashion. If you reboot the card, it will revert back to "disabled" but v120 will still be able to connect. Secondly, there is a known problem with getting Cisco ISDN adapters to connect. I'm currently hammering this through 3Com support as ticket #34184 V110 Rate Adaption: disable Force Fixed Network Rate: notForced Force Network Rate Speed: kbps56 Enable 45-65 Second Link: noDelay Analog Calls Over Digital: allow Async PPP/Sync PPP Conversion: disable X.75: disable Set Data Mode of Modem: autodetect Set Originate HDLC Protocol: v120 Set Originate Non-HDLC Protocol: none Set Originate Analog Modem/Fax: analogModemFax V120: enable X75 Frame Size: 2048 X75 Window Size: 2 >- One HDM got stuck twice (I had to "hardware reset" it from the >TCM menu), and a HiPerARC once for no reason at all. Syslog gave me no >clue at all. Any opinions on this will be welcome, as I'd hate to nuke >200 users from a fully loaded chassis, even with two ARC's. Is there a >way to have one spare ARC so the user will be moved from one to the >other by some magical feature that USR implemented ? I've had this behavior happen when I move PRIs around, or one becomes unplugged and needs to be plugged back in. The HDM has a hard time recovering from an interruption of the entire circuit. Check your wiring, and talk to your telco about the PRIs stability.

James Wilson said once upon a time: > >In the never ending drama of setting up our first tc hub... > >Well, we finaly can connect to the thing. ICG has set the PRI to send 7 >digits on the D channel instead of 0, and now there are no more fast busies. > However, we can only connect at 9600 baud. What do I need to check to get >this working right? TCM -> Modem Settings -> Call Control Options -> DNIS Based Incoming Call Digits = 7 Set the ANI as well. I had this exact same problem when I set up my first DSS circuit. Your modems are answering to early because the phone co is still sending digits, thus the other side doesn't get vital negotiation tones.

Curt Shambeau said once upon a time: >Plugged in a standard 1866/1706 chassis with 1 70A PS - 48 ports with Quad >modems. The Load-o-meter(tm) went up to 2 lights. > >Plugged in my fully loaded chassis - turned on 1 of the 70A power supplies >- the load-o-meter went up to 2 lights - turned on the 2nd 70A power >supply on the fully loaded chassis, the load-o-meter went up to 3 lights. > >My guess is that a fully loaded HiPer chassis takes just slightly more >power than a standard 48-port quad modem chassis. I wonder if power usage goes up when it is loaded with calls?

Tatai SV Krishnan said once upon a time: >Each HiPer ARC with the current code supports 7 HDM - you can have two >HiPer ARC in one chassis, but make sure that you disable chassis awarness >on the HiPer ARCS. "disable nmc chas " >Also you must configure each HiPer ARC to be owner of only 7 HDM. Why am I running 10? Or does this apply to chassis with two ARCs?

Brian said once upon a time: >However, if a user is assigned a Framed-Route by RADIUS, which is not in >an ippool on the HyperARC, then it would appear that the HyperARC does not >send out a RIP update, because my UNIX servers, running gated(ripv2), >don't see anything. If a user with: > >Framed-Route "208.214.45.0/28 208.214.45.1 1" Brian, you might be affected by the same bug I was having trouble with. Get the 4.0.89 engineering release from 3Com. Feel free to reference my ticket on the problem, #27271.

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

Brian said once upon a time: >I heard there was a way to access the HyperDSP's from within the HyperARC, >anyone care to comment on this? Forgive me if the question is a simple >one, but I have only been running this HyperARC for a few hours and I am >very anxious to learn as much as I can. You can do hangups and look at the speed of the connects via the interfaces: hangup int (interface) show int (interface) >Our plan is to slam chassis with as many HyperDSP's as they will take, and >put dual HyperARC's in them, but what I am wondering, is it really >necessary or is performance hit marginal? I'm running with 10 in one chassis right now with one ARC. My Quake performance has been extremely good, even when it is stacked with 230 callers. I think I might just stick to one ARC per chassis, at least until the cost comes down.

Brian said once upon a time: >At 19:12:54, Facility "Call Initiation Process", Leve "CRITICAL":: CIP >Set the type to dial_out in the user record I think it is a bogus RADIUS interpretation. It happens whenever I have a shell user login. Ignore it.

>Putting together our first fully loaded HiPer Chassis. However, the 130A >power supplies are on back order. Loaded up a 70A chassis with 14 HDM >cards and 2 HiPer ARC cards. No problem for even 1 70A power supply as >of yet... > >I'll try and do some more real world load readings ASAP. Very crude testing: I have a APC SU1400 that has some load on it already - 1 light out of 5 on the load-o-meter. Plugged in a standard 1866/1706 chassis with 1 70A PS - 48 ports with Quad modems. The Load-o-meter(tm) went up to 2 lights. Plugged in my fully loaded chassis - turned on 1 of the 70A power supplies - the load-o-meter went up to 2 lights - turned on the 2nd 70A power supply on the fully loaded chassis, the load-o-meter went up to 3 lights. My guess is that a fully loaded HiPer chassis takes just slightly more power than a standard 48-port quad modem chassis. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

Here's the hypothetical scenario. Customer with Livingston ISDN OR, I have Netserver 8/I-modem. The 8 B channels connected to my Netserver are in a hunt group with one number for customers to dial. Dial on demand initiated by OR to one of my B channels on Netserver. (authenticated with radius and static IP)...customer has 64K ISDN for access to the Internet. Is it possible to set up the Office Router on customer end, and Netserver on my end, such that in times of heavy traffic the OR can automatically dial a second B channel to get 128K (not necessarily on the same BRI on my end as they are in a pool) ...another ISP is offering this prospective customer a slightly cheaper solution (intital and monthly) using a 486 running Linux, with ISDN adapter...dial on demand, and IP masquerading. I personally think that the Livingston OR would be more stable, and require less maintenance. Any thoughts on what could be done here? Curt

> Verrrry interesting. I wonder what the 130A power supplys are all about > then? I had heard they engineered and started production on them, in anticipation of needing more power, but in the end they didn't need any more power. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

> >My guess is that a fully loaded HiPer chassis takes just slightly more > >power than a standard 48-port quad modem chassis. > > I wonder if power usage goes up when it is loaded with calls? I'm sure it will, but only time will tell. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

I dug up some glossies from our USR file and found power consumption for everything except the NMC card. I would imagine that it is comparable to the Netserver card. Here's what it states: Quad Modem: 1.5A @ 5VDC, 0.1A @ 12VDC, 8.7 watts Netserver: 5 watts Dual PRI: 5 watts NMC card: 5 watts? (a guess) HiPer DSP (24 channel): 4.8A typical (24 watts) HiPer ARC: 6.2A typical (31 watts) Now, presuming my recollection of physics is correct, amps * volts = watts. This would mean a 70 AMP power supply on 110 volts = 7700 watts. However, I don't think this is correct because there is a lot of loss in the conversion. A fully loaded chassis with 14 DSPs (336 watts), two ARCs (61 watts), and an NMC card (5 watts?) would be 402 watts. A recommended chassis with 10 DSPs (240 watts), two ARCs (61 watts), and an NMC card (5 watts) would be 306 watts. A fully loaded chassis with 12 Quad modems (105 watts), one Netserver (5 watts), one NMC (5 watts), and one PRI card (5 watts) consumes only 120 watts in comparison. The real question is what is the output of the power supply. I don't think it is 7700 watts. That would be overkill, plus nuking-from-orbit. I couldn't find any docomentation on the 70 AMP power supply at all. Anyone else? If you go off the AMP ratings stated only on the DSP and the ARC glossies, then USR's recommendations are inline.

yes. we have a matrix 5000xr ups just for our total control hubs and it has a neat little readout that tells you what % of the ups' power is being drawn out. When we have a sudden rush of people logging in or compared between the middle of the night and prime time you can see a definite difference in the % of power being used. when we added redundant power supplies we saw a small % rise in the consumption. matthew At 01:34 PM 1/28/98 -0700, Pete Ashdown wrote: >Curt Shambeau said once upon a time: > >>Plugged in a standard 1866/1706 chassis with 1 70A PS - 48 ports with Quad >>modems. The Load-o-meter(tm) went up to 2 lights. >> >>Plugged in my fully loaded chassis - turned on 1 of the 70A power supplies >>- the load-o-meter went up to 2 lights - turned on the 2nd 70A power >>supply on the fully loaded chassis, the load-o-meter went up to 3 lights. >> >>My guess is that a fully loaded HiPer chassis takes just slightly more >>power than a standard 48-port quad modem chassis. > >I wonder if power usage goes up when it is loaded with calls? > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

yes. we have a matrix 5000xr ups just for our total control hubs and it has a neat little readout that tells you what % of the ups' power is being drawn out. When we have a sudden rush of people logging in or compared between the middle of the night and prime time you can see a definite difference in the % of power being used. when we added redundant power supplies we saw a small % rise in the consumption. matthew At 01:34 PM 1/28/98 -0700, Pete Ashdown wrote: >Curt Shambeau said once upon a time: > >>Plugged in a standard 1866/1706 chassis with 1 70A PS - 48 ports with Quad >>modems. The Load-o-meter(tm) went up to 2 lights. >> >>Plugged in my fully loaded chassis - turned on 1 of the 70A power supplies >>- the load-o-meter went up to 2 lights - turned on the 2nd 70A power >>supply on the fully loaded chassis, the load-o-meter went up to 3 lights. >> >>My guess is that a fully loaded HiPer chassis takes just slightly more >>power than a standard 48-port quad modem chassis. > >I wonder if power usage goes up when it is loaded with calls? > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

> Now, presuming my recollection of physics is correct, amps * volts = watts. > This would mean a 70 AMP power supply on 110 volts = 7700 watts. However, > I don't think this is correct because there is a lot of loss in the > conversion. I'm sure the 70AMP rating is DC AMPS, not dependant on the 110V AC. So it probably puts out like 60A of 5V and 10A of 12v, or something like that. It's hard to say. I'd have to look up the documentation to see if that is true or not. However, this case would yeild 420 watts. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

> Now, presuming my recollection of physics is correct, amps * volts = watts. > This would mean a 70 AMP power supply on 110 volts = 7700 watts. However, > I don't think this is correct because there is a lot of loss in the > conversion. Found it... 5.2vDC 70 Amps -5v DC 2 Amps 12.2v DC 5.5 Amps -12.2v DC 5.5 Amps Thus, the rating on the power supply is ONLY the +5v signal. In the docs, it rates the 70A power supply at 670 watts. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

I installed my HDM's and a HiPerARC. Now I've got two major problems : - ISDN does connect with lots of difficulties, I had to really fiddle with my trusty Elite2864i (Zyxel) and it connects OK now, but I can't expect this from my clients. I snooped one the connections using the "monitor PPP" feature from the HiPerARC, and got some weird tracing messages (I'll send anyone interested this logfile). - One HDM got stuck twice (I had to "hardware reset" it from the TCM menu), and a HiPerARC once for no reason at all. Syslog gave me no clue at all. Any opinions on this will be welcome, as I'd hate to nuke 200 users from a fully loaded chassis, even with two ARC's. Is there a way to have one spare ARC so the user will be moved from one to the other by some magical feature that USR implemented ? My opinion is that the HDM / ARC combo looks to be a killer, but needs some work with connection problems with some "el-cheapo" modems like the Rockwell RC288DPi based-modems, which will connect at 28800 but not at 33600 reliably. Thanks for any replies, Robert von Bismarck Petrel Communication S.A.

We just upgraded our system from 1 PRI to 2 PRI's. When we did this, our ISDN connections went to busy signals even though there are plenty of open ports. Does anyone have the configuration for the Total Control Hub to verify that all the ISDN is working correctly. Any help or advice would be appreaciated. 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

-> I dug up some glossies from our USR file and found power consumption for -> everything except the NMC card. I would imagine that it is comparable to -> the Netserver card. Here's what it states: -> -> Quad Modem: 1.5A @ 5VDC, 0.1A @ 12VDC, 8.7 watts -> Netserver: 5 watts -> Dual PRI: 5 watts -> NMC card: 5 watts? (a guess) -> HiPer DSP (24 channel): 4.8A typical (24 watts) -> HiPer ARC: 6.2A typical (31 watts) I suspect it is 70A at 5VDC or 350 watts. 7700 watts is a lot of current and heat. Jeff Binkley ASA Network Computing

Is there a way to see where a call orginates from on the TC hubs? In the logs or can I telnet into the box to check? I think there was a discussion about this earlier but I can't remember the results. We are running channelized T1's from USWorst. 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 ~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~

On Thu, 29 Jan 1998, Mark van Wouw wrote: > At 05:27 AM 1/28/98 -0600, Tatai SV Krishnan wrote: > >On Wed, 28 Jan 1998, Pete Ashdown wrote: > > > >> Tatai SV Krishnan said once upon a time: > >> > >> >Each HiPer ARC with the current code supports 7 HDM - you can have two > >> >HiPer ARC in one chassis, but make sure that you disable chassis awarness > >> >on the HiPer ARCS. "disable nmc chas " > >> >Also you must configure each HiPer ARC to be owner of only 7 HDM. > >> > >> Why am I running 10? Or does this apply to chassis with two ARCs? > > > >The HiPer ARC is capable of having 10 HDM in a chassis. The current > >release of code - we suggest users to upto 7 modems. Also you can have 2 > >HiPer ARC in one chassis. > > > > This is different from the original specs I heard on the HiPer ARC. From > a message early in the beta testing: > > The HiPer Access Router Card is a RISC based NAC that can support up > to 14 HiPer DSP NACs in one chassis. It is our next generation Router > card for supporting high call density and high call volume. It can > also be used with the older Quad modem cards. The recommended maximum > configuration will have 14 HiPer DSPs, and 2 HiPer ARCs, which will > allow for redundancy or load sharing. > I had just said the same as above, the HiPer ARC is capable of 14 DSP, We do recommend to have two HiPer ARC for 14 DSPs. krish > I believe your comments on this list are the first I've seen, from an > official USR/3-Com source that are contrary to the above...but maybe > I missed something. > > 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. > \ 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

A lot of you are interested in OSPF. I would like to start a discussion on this subject. Namely I would like to know from you the following a. What type of features that you would like to see in TC supporting OSPF b. Any specific authentication types c. Your comments on how your networks are setup and the products/problems that you see in your current installation. 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 Thu, 29 Jan 1998, Jeff Binkley wrote: > > > > U>As you know we installed a hiperarc yesterday with few problems. At > U>first, we forgot to turn off MPIP which screwed up ISDN PPP > U>negotiations. And a few subnets are problematic for which we have an > U>update that fixes framed-routes. > U>However we are experiencing a new problem that suddenly affects about > U>25% of our users.. And we can fix it about 75% of the time.. It > U>corresponded to the HiperARC upgrade but affects all chassis.. HiPER ARC and MPIP? HiPer ARC currently does not support MPIP. What is the problem exactly? ARe you talking about MPIP or HiPerDSP? > > Can someone give a 2 minute explanation on why MPIP affects ISDN PPP > connections and if that also holds true to Netservers ? NETServer supports MPIP, The NETServer doing MPIP should have the same version of code. One NETServer should be the master and client the other netservers should be the client. The secret should match and also the time server should be set on all NETServer. If this is not done then you will have problems only on MPIP calls not on all ISDN calls. krish > > Jeff Binkley > ASA Network Computing > > CMPQwk 1.42 9999 > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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

We use ICG in Cleveland for our PRI's and the first one has been OK for 8 months. we have seen 52 & 53K connects. Most in the mid 40's though. The verdic is still out on the second PRI. I have never seen the work order to see how they provisioned it, (it was provisioned by a third party) but I will ask them. ICG has been very good to us. Im sure it's the local tech at the switch that's in error. PS key to get them to look quickly at the problem is to have your circut ID handy not your phone # When I get the settings that work for us i'll them to e-mail you. Mike Hamrich DrFast.Net, Inc.

When you have problems with X2 ask your clients to connect to the chassis using a terminal program, use the escape code +++ after connection and get the following information ati4 ati6 ati11 aty11 This information will allow 3Com support to find out where the problem is. If the customer can dial and connect X2 to 3com/USR site and if he fails to connect to your site then have the above information for both the connections. When you give this information it will help us to indentify and solve your problem. krish On Thu, 29 Jan 1998, Brian Elfert wrote: > I have a couple of users who have Sportster 840s that were chip upgraded > from 33.6 to x2. > > Neither of them can connect to a USR TC rack at x2 speeds. They just get > V.34 connects. One of them can connect to a USR MP/8I at x2 speeds just > fine. The other one we haven't tried with the MP/8I yet. > > I noticed some people with the same modem model complaining that they > don't get x2 connects in the comp.dcom.modems newsgroup. > > Are there any known problems with the chip upgraded Sportster 840 > connecting at x2 speed? Anyone else seeing this problem? I called USR, > and they said to call the Sportster tech support. > > 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

As you know we installed a hiperarc yesterday with few problems. At first, we forgot to turn off MPIP which screwed up ISDN PPP negotiations. And a few subnets are problematic for which we have an update that fixes framed-routes. However we are experiencing a new problem that suddenly affects about 25% of our users.. And we can fix it about 75% of the time.. It corresponded to the HiperARC upgrade but affects all chassis.. PROBLEM: Some users can't connect.. They receive the following error.. This happens on both arc equipped and netserver equipped chassis.. "Dial-up Networking could not negotiate a compatible set of network protocols you specified in the Server Type settings. Check you Network Configuration in the Control Panel then try the connection again." And one user reported this error: "The computer you are dialing into does not respond to a network request. Check your server type settings in the properties of the connection. If the problem persists, check with your network administrator" What Radius sees during a successful connection: Netserver: rad_authentication: xxx/xxx 'user' at 208.206.76.37 PPP Authentication: xxx/xxx 'user' via 208.206.76.37 from usr3ts1 port xxx PPP-OK Accounting: xxx/xxx 'user' via 208.206.76.37 from usr3ts1 port xxx Start-OK HiperARC: rad_authentication: xxx/xxx 'user' at 208.206.76.35 PPP dumb Authentication: xxx/xxx 'user' via 208.206.76.35 from 208.206.76.35 port xxx PPP-OK Accounting: xxx/xxx 'user' via 208.206.76.35 from 208.206.76.35 port xxx Start-OK TCM shows the following differences between an ARC handled call and one handled by a netserver.. All other parms are the same... Parm Netserver HiperARC Number of Rings before DTR 1 0 HST Speed Back Channel Speed bps300(2) bps450(1) B Channel Used for the Call 7 0 Timing Offset in parts per million 65522 58256 Carrier Offset in Hertz 65535 876 When we have a user that is experiencing the problem, we can usually fix it by deleting his dialup network adapter and reinstalling it from scratch.. This works most of the time but not all.. Also repeated attempts (10 or 20) sometimes results in a successful connection.. Radius logs no information when user logs on and receives this "unable to negotiate network protocol" error.. Brian is out-of-town on a family emergency and my tech staff is going nutz. (our tech support supervisior is also his girlfriend and went with him).. It's like having both my hands tied.. Any of you have a clue as to what might be the problem here. If there is anything else of help that I can post, please let me know.. Thanks 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

I am using the release version of the HyperARC code. I am also using the latest version of Merit RADIUS. All of the users who are authenticated and log on the chassis with the Hyperarc in it show "dumb" as there login type, is this right? /-------------------------- 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 -----------------------------/

We are trying to configure a brand new USR Netserver I modem plus 16. We upgraded it to modem code 2.2.2 (x2). We are a small ISP trying to use this box to serve ISDN and X2. We are using Merit Radius (not the paid one). We are having two problems right now. 1. It isn't passing DNS to the callers. I believe at one point it actually did but, now it is NOT doing this. What this means is that any of our callers trying to use this server for X2 or ISDN have to specify at least primary DNS in their WIN95 DUN icon for us. Called Totalservice Support and the tech said that this unit can NOT pass DNS server info to the caller. Is this true? And if so why? I have 3 year old Chase termservers (another story) that can do this. 2. When we issue show interface mod:x (x being the port) it doesn't report back ANY username and the speed field reports the wrong connect speed. I connected several times between 44K to 47K but, the server said it was 33333 for all of them? We tried a local user and a radius user and this seems to have no effect on this at all. Any ideas would be most welcome.

At 07:59 PM 1/28/98 -0700, Greg Coffey wrote: >Is there a way to see where a call orginates from on the TC hubs? In the >logs or can I telnet into the box to check? I think there was a discussion >about this earlier but I can't remember the results. We are running >channelized T1's from USWorst. > check the Performance Monitor in TCM.. Select a slot/channel and the parameter "Number of Last Caller (ANI)".. The service must also be provisioned right from the telco.. 0, 3, 7, or 10 digits, etc.. allen >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 >~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~ > > >- > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. > >

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

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-encapsulated 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

U>As you know we installed a hiperarc yesterday with few problems. At U>first, we forgot to turn off MPIP which screwed up ISDN PPP U>negotiations. And a few subnets are problematic for which we have an U>update that fixes framed-routes. U>However we are experiencing a new problem that suddenly affects about U>25% of our users.. And we can fix it about 75% of the time.. It U>corresponded to the HiperARC upgrade but affects all chassis.. Can someone give a 2 minute explanation on why MPIP affects ISDN PPP connections and if that also holds true to Netservers ? Jeff Binkley ASA Network Computing CMPQwk 1.42 9999

Hi I am running a small ISP, one of my customer is using linux to connect to TC, and claim he can not ping my end after connected. FOr the win95 users, everything is okay. I review the client's ppplog file and his routeing table, it seems to me, everthing is fine, and I can ping back that user after the connection. THe client just chaange the phone number, username and password then can connect to another ISP without an problem, so I wonder is anyone there can point me a direction ? Thank you! the following is client's ppp log file Jan 27 23:15:23 center pppd[500]: pppd 2.3.3 started by root, uid 0 Jan 27 23:15:43 center pppd[500]: Serial connection established. Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set flags to 70000 Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set flags to 70000 Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set xasyncmap Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set xmit asyncmap ffffffff Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set flags to 70000 Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set mru to 5dc Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set rcv asyncmap ffffffff Jan 27 23:15:44 center kernel: ppp_tty_ioctl: set flags to 70000 Jan 27 23:15:44 center kernel: ppp: write frame, count = 18 Jan 27 23:15:44 center kernel: FF 03 C0 21 01 01 00 0E ...!.... Jan 27 23:15:44 center kernel: 05 06 DD 47 25 0A 07 02 ...G%... Jan 27 23:15:44 center kernel: 08 02 .. Jan 27 23:15:44 center pppd[500]: Using interface ppp0 Jan 27 23:15:44 center pppd[500]: Connect: ppp0 <--> /dev/ttyS0 Jan 27 23:15:44 center pppd[500]: sent [LCP ConfReq id=0x1 <magic 0xdd47250a> <pcomp> <accomp>] Jan 27 23:15:47 center kernel: ppp: write frame, count = 18 Jan 27 23:15:47 center kernel: FF 03 C0 21 01 01 00 0E ...!.... Jan 27 23:15:47 center kernel: 05 06 DD 47 25 0A 07 02 ...G%... Jan 27 23:15:47 center kernel: 08 02 .. Jan 27 23:15:47 center pppd[500]: sent [LCP ConfReq id=0x1 <magic 0xdd47250a> <pcomp> <accomp>] Jan 27 23:15:47 center kernel: ppp: frame with bad fcs, length = 66 Jan 27 23:15:47 center kernel: ppp: bad frame, count = 66 Jan 27 23:15:47 center kernel: FF 03 20 03 50 50 50 20 .. .PPP Jan 27 23:15:47 center kernel: 73 65 73 73 69 6F 6E 20 session Jan 27 23:15:47 center kernel: 66 72 6F 6D 20 28 32 30 from (20 Jan 27 23:15:47 center kernel: 34 2E 32 34 34 2E 31 30 4.244.10 Jan 27 23:15:47 center kernel: 31 2E 32 35 30 29 20 74 1.250) t Jan 27 23:15:47 center kernel: 6F 20 4E 65 67 6F 74 69 o Negoti Jan 27 23:15:47 center kernel: 61 74 65 64 20 62 65 67 ated beg Jan 27 23:15:47 center kernel: 69 6E 6E 69 6E 67 2E 2E inning.. Jan 27 23:15:47 center kernel: 2E 2E .. Jan 27 23:15:47 center kernel: ppp: receive frame, count = 28 Jan 27 23:15:47 center kernel: FF 03 C0 21 01 01 00 18 ...!.... Jan 27 23:15:47 center kernel: 02 06 00 00 00 00 05 06 ........ Jan 27 23:15:47 center kernel: DB 28 43 A8 07 02 08 02 .(C..... Jan 27 23:15:47 center kernel: 03 04 C0 23 ...# Jan 27 23:15:47 center kernel: ppp: write frame, count = 28 Jan 27 23:15:47 center pppd[500]: rcvd [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0xdb2843a8> <pcomp> <accomp> <auth pap>] Jan 27 23:15:47 center pppd[500]: sent [LCP ConfAck id=0x1 <asyncmap 0x0> <magic 0xdb2843a8> <pcomp> <accomp> <auth pap>] Jan 27 23:15:47 center kernel: FF 03 C0 21 02 01 00 18 ...!.... Jan 27 23:15:47 center kernel: 02 06 00 00 00 00 05 06 ........ Jan 27 23:15:47 center kernel: DB 28 43 A8 07 02 08 02 .(C..... Jan 27 23:15:47 center kernel: 03 04 C0 23 ...# Jan 27 23:15:50 center kernel: ppp: write frame, count = 18 Jan 27 23:15:50 center kernel: FF 03 C0 21 01 01 00 0E ...!.... Jan 27 23:15:50 center kernel: 05 06 DD 47 25 0A 07 02 ...G%... Jan 27 23:15:50 center kernel: 08 02 .. Jan 27 23:15:50 center pppd[500]: sent [LCP ConfReq id=0x1 <magic 0xdd47250a> <pcomp> <accomp>] Jan 27 23:15:50 center pppd[500]: rcvd [LCP ConfAck id=0x1 <magic 0xdd47250a> <pcomp> <accomp>] Jan 27 23:15:50 center pppd[500]: sent [PAP AuthReq id=0x1 user="test" password="test"] Jan 27 23:15:50 center kernel: ppp: receive frame, count = 18 Jan 27 23:15:50 center kernel: FF 03 C0 21 02 01 00 0E ...!.... Jan 27 23:15:50 center kernel: 05 06 DD 47 25 0A 07 02 ...G%... Jan 27 23:15:50 center kernel: 08 02 .. Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set xmit asyncmap 0 Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set flags to f070003 Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set mru to 5dc Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set rcv asyncmap ffffffff Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set flags to f070003 Jan 27 23:15:50 center kernel: ppp: write frame, count = 26 Jan 27 23:15:50 center kernel: FF 03 C0 23 01 01 00 16 ...#.... Jan 27 23:15:50 center kernel: 06 77 65 73 6C 61 6E 0A .test. Jan 27 23:15:50 center kernel: 77 65 73 6C 65 79 31 39 tes Jan 27 23:15:50 center kernel: 39 38 98 Jan 27 23:15:50 center kernel: ppp: receive frame, count = 24 Jan 27 23:15:50 center kernel: FF 03 C0 23 02 01 00 14 ...#.... Jan 27 23:15:50 center kernel: 0F 4C 6F 67 69 6E 20 53 .Login S Jan 27 23:15:50 center pppd[500]: rcvd [PAP AuthAck id=0x1 "Login Succeeded"] Jan 27 23:15:50 center pppd[500]: Remote message: Login Succeeded Jan 27 23:15:50 center pppd[500]: sent [IPCP ConfReq id=0x1 <addr 204.244.101.1> <compress VJ 0f 01>] Jan 27 23:15:50 center pppd[500]: rcvd [IPCP ConfReq id=0x1 <compress VJ 0f 00> <addr 204.244.101.250>] Jan 27 23:15:50 center pppd[500]: sent [IPCP ConfAck id=0x1 <compress VJ 0f 00> <addr 204.244.101.250>] Jan 27 23:15:50 center pppd[500]: rcvd [IPCP ConfAck id=0x1 <addr 204.244.101.1> <compress VJ 0f 01>] Jan 27 23:15:50 center pppd[500]: local IP address 204.244.101.1 Jan 27 23:15:50 center pppd[500]: remote IP address 204.244.101.250 Jan 27 23:15:50 center pppd[500]: rcvd [CCP ConfReq id=0x1 < 12 06 00 00 00 01> < 11 05 00 01 04>] Jan 27 23:15:50 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:15:50 center pppd[500]: sent [CCP ConfRej id=0x1 < 12 06 00 00 00 01> < 11 05 00 01 04>] Jan 27 23:15:50 center kernel: 75 63 63 65 65 64 65 64 ucceeded Jan 27 23:15:50 center kernel: ppp: receive frame, count = 20 Jan 27 23:15:50 center kernel: FF 03 80 21 01 01 00 10 ...!.... Jan 27 23:15:50 center kernel: 02 06 00 2D 0F 00 03 06 ...-.... Jan 27 23:15:50 center kernel: CC F4 65 FA ..e. Jan 27 23:15:50 center kernel: ppp: write frame, count = 20 Jan 27 23:15:50 center kernel: FF 03 80 21 01 01 00 10 ...!.... Jan 27 23:15:50 center kernel: 03 06 CC F4 65 01 02 06 ....e... Jan 27 23:15:50 center kernel: 00 2D 0F 01 .-.. Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set flags to f070043 Jan 27 23:15:50 center kernel: ppp: write frame, count = 20 Jan 27 23:15:50 center kernel: FF 03 80 21 02 01 00 10 ...!.... Jan 27 23:15:50 center kernel: 02 06 00 2D 0F 00 03 06 ...-.... Jan 27 23:15:50 center kernel: CC F4 65 FA ..e. Jan 27 23:15:50 center kernel: ppp: receive frame, count = 20 Jan 27 23:15:50 center kernel: FF 03 80 21 02 01 00 10 ...!.... Jan 27 23:15:50 center kernel: 03 06 CC F4 65 01 02 06 ....e... Jan 27 23:15:50 center kernel: 00 2D 0F 01 .-.. Jan 27 23:15:50 center kernel: ppp: receive frame, count = 19 Jan 27 23:15:50 center kernel: FF 03 80 FD 01 01 00 0F ........ Jan 27 23:15:50 center kernel: 12 06 00 00 00 01 11 05 ........ Jan 27 23:15:50 center kernel: 00 01 04 ... Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set maxcid to 16 Jan 27 23:15:50 center kernel: ppp_tty_ioctl: set flags to f07004f Jan 27 23:15:50 center kernel: ppp: channel ppp0 going up for IP packets! Jan 27 23:15:50 center kernel: ppp: set np 0 to 0 Jan 27 23:15:50 center kernel: ppp: write frame, count = 8 Jan 27 23:15:50 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:15:50 center kernel: ppp: write frame, count = 19 Jan 27 23:15:50 center kernel: FF 03 80 FD 04 01 00 0F ........ Jan 27 23:15:50 center kernel: 12 06 00 00 00 01 11 05 ........ Jan 27 23:15:50 center kernel: 00 01 04 ... Jan 27 23:15:53 center kernel: ppp: receive frame, count = 19 Jan 27 23:15:53 center kernel: FF 03 80 FD 01 02 00 0F ........ Jan 27 23:15:53 center kernel: 12 06 00 00 00 01 11 05 ........ Jan 27 23:15:53 center kernel: 00 01 04 ... Jan 27 23:15:53 center kernel: ppp: write frame, count = 19 Jan 27 23:15:53 center kernel: FF 03 80 FD 04 02 00 0F ........ Jan 27 23:15:53 center kernel: 12 06 00 00 00 01 11 05 ........ Jan 27 23:15:53 center kernel: 00 01 04 ... Jan 27 23:15:54 center kernel: ppp: write frame, count = 8 Jan 27 23:15:54 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:15:53 center pppd[500]: rcvd [CCP ConfReq id=0x2 < 12 06 00 00 00 01> < 11 05 00 01 04>] Jan 27 23:15:53 center pppd[500]: sent [CCP ConfRej id=0x2 < 12 06 00 00 00 01> < 11 05 00 01 04>] Jan 27 23:15:54 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:15:56 center kernel: ppp: receive frame, count = 19 Jan 27 23:15:56 center kernel: FF 03 80 FD 01 03 00 0F ........ Jan 27 23:15:56 center kernel: 12 06 00 00 00 01 11 05 ........ Jan 27 23:15:56 center kernel: 00 01 04 ... Jan 27 23:15:56 center kernel: ppp: write frame, count = 19 Jan 27 23:15:56 center kernel: FF 03 80 FD 04 03 00 0F ........ Jan 27 23:15:56 center kernel: 12 06 00 00 00 01 11 05 ........ Jan 27 23:15:56 center kernel: 00 01 04 ... Jan 27 23:15:56 center pppd[500]: rcvd [CCP ConfReq id=0x3 < 12 06 00 00 00 01> < 11 05 00 01 04>] Jan 27 23:15:56 center pppd[500]: sent [CCP ConfRej id=0x3 < 12 06 00 00 00 01> < 11 05 00 01 04>] Jan 27 23:15:57 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:15:57 center kernel: ppp: write frame, count = 8 Jan 27 23:15:57 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:00 center kernel: ppp: write frame, count = 8 Jan 27 23:16:00 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:00 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:03 center kernel: ppp: write frame, count = 8 Jan 27 23:16:03 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:03 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:06 center kernel: ppp: write frame, count = 8 Jan 27 23:16:06 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:06 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:09 center kernel: ppp: write frame, count = 8 Jan 27 23:16:09 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:09 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:12 center kernel: ppp: write frame, count = 8 Jan 27 23:16:12 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:12 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:15 center kernel: ppp: write frame, count = 8 Jan 27 23:16:15 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:15 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:18 center kernel: ppp: write frame, count = 8 Jan 27 23:16:18 center kernel: FF 03 80 FD 01 01 00 04 ........ Jan 27 23:16:18 center pppd[500]: sent [CCP ConfReq id=0x1] Jan 27 23:16:21 center pppd[500]: CCP: timeout sending Config-Requests --> timeout here Jan 27 23:16:23 center kernel: ppp: receive frame, count = 56 Jan 27 23:16:23 center kernel: FF 03 00 21 45 00 00 34 ...!E..4 Jan 27 23:16:23 center kernel: 0A 89 00 00 FF 11 4C 4B ......LK Jan 27 23:16:23 center kernel: CC F4 65 FA CC F4 65 01 ..e...e. Jan 27 23:16:23 center kernel: 02 08 02 08 00 20 62 C0 ..... b. Jan 27 23:16:23 center kernel: 02 01 00 00 00 02 00 00 ........ Jan 27 23:16:23 center kernel: CC F4 65 00 00 00 00 00 ..e..... Jan 27 23:16:23 center kernel: 00 00 00 00 00 00 00 01 ........ Jan 27 23:16:23 center kernel: ppp: write frame, count = 84 Jan 27 23:16:23 center kernel: FF 03 00 21 45 C0 00 50 ...!E..P Jan 27 23:16:23 center kernel: 00 B2 00 00 40 01 14 57 ....@..W Jan 27 23:16:23 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:16:23 center kernel: 03 03 62 13 00 00 00 00 ..b..... Jan 27 23:16:23 center kernel: 45 00 00 34 0A 89 00 00 E..4.... Jan 27 23:16:23 center kernel: FF 11 4C 4B CC F4 65 FA ..LK..e. Jan 27 23:16:23 center kernel: CC F4 65 01 02 08 02 08 ..e..... Jan 27 23:16:23 center kernel: 00 20 62 C0 02 01 00 00 . b..... Jan 27 23:16:23 center kernel: 00 02 00 00 CC F4 65 00 ......e. Jan 27 23:16:23 center kernel: 00 00 00 00 00 00 00 00 ........ Jan 27 23:16:23 center kernel: 00 00 00 01 .... Jan 27 23:16:55 center kernel: ppp: receive frame, count = 56 Jan 27 23:16:55 center kernel: FF 03 00 21 45 00 00 34 ...!E..4 Jan 27 23:16:55 center kernel: 0A 8B 00 00 FF 11 4C 49 ......LI Jan 27 23:16:55 center kernel: CC F4 65 FA CC F4 65 01 ..e...e. Jan 27 23:16:55 center kernel: 02 08 02 08 00 20 62 C0 ..... b. Jan 27 23:16:55 center kernel: 02 01 00 00 00 02 00 00 ........ Jan 27 23:16:55 center kernel: CC F4 65 00 00 00 00 00 ..e..... Jan 27 23:16:55 center kernel: 00 00 00 00 00 00 00 01 ........ Jan 27 23:16:55 center kernel: ppp: write frame, count = 84 Jan 27 23:16:55 center kernel: FF 03 00 21 45 C0 00 50 ...!E..P Jan 27 23:16:55 center kernel: 00 B3 00 00 40 01 14 56 ....@..V Jan 27 23:16:55 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:16:55 center kernel: 03 03 62 13 00 00 00 00 ..b..... Jan 27 23:16:55 center kernel: 45 00 00 34 0A 8B 00 00 E..4.... Jan 27 23:16:55 center kernel: FF 11 4C 49 CC F4 65 FA ..LI..e. Jan 27 23:16:55 center kernel: CC F4 65 01 02 08 02 08 ..e..... Jan 27 23:16:55 center kernel: 00 20 62 C0 02 01 00 00 . b..... Jan 27 23:16:55 center kernel: 00 02 00 00 CC F4 65 00 ......e. Jan 27 23:16:55 center kernel: 00 00 00 00 00 00 00 00 ........ Jan 27 23:16:55 center kernel: 00 00 00 01 .... Jan 27 23:17:11 center kernel: ppp: write frame, count = 88 Jan 27 23:17:11 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:11 center kernel: 00 B4 00 00 40 01 15 11 ....@... Jan 27 23:17:11 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:11 center kernel: 08 00 ED F3 04 02 00 00 ........ Jan 27 23:17:11 center kernel: F7 6A CE 34 48 67 0D 00 .j.4Hg.. Jan 27 23:17:11 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:11 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:11 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:11 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:11 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:11 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:12 center kernel: ppp: write frame, count = 88 Jan 27 23:17:12 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:12 center kernel: 00 B5 00 00 40 01 15 10 ....@... Jan 27 23:17:12 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:12 center kernel: 08 00 C1 49 04 02 01 00 ...I.... Jan 27 23:17:12 center kernel: F8 6A CE 34 72 11 0E 00 .j.4r... Jan 27 23:17:12 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:12 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:12 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:12 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:12 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:12 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:13 center kernel: ppp: write frame, count = 88 Jan 27 23:17:13 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:13 center kernel: 00 B6 00 00 40 01 15 0F ....@... Jan 27 23:17:13 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:13 center kernel: 08 00 3C 51 04 02 02 00 ..<Q.... Jan 27 23:17:13 center kernel: F9 6A CE 34 F5 09 0E 00 .j.4.... Jan 27 23:17:13 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:13 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:13 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:13 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:13 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:13 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:14 center kernel: ppp: write frame, count = 88 Jan 27 23:17:14 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:14 center kernel: 00 B7 00 00 40 01 15 0E ....@... Jan 27 23:17:14 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:14 center kernel: 08 00 C0 49 04 02 03 00 ...I.... Jan 27 23:17:14 center kernel: FA 6A CE 34 6F 11 0E 00 .j.4o... Jan 27 23:17:14 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:14 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:14 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:14 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:14 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:14 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:15 center kernel: ppp: write frame, count = 88 Jan 27 23:17:15 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:15 center kernel: 00 B8 00 00 40 01 15 0D ....@... Jan 27 23:17:15 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:15 center kernel: 08 00 36 51 04 02 04 00 ..6Q.... Jan 27 23:17:15 center kernel: FB 6A CE 34 F7 09 0E 00 .j.4.... Jan 27 23:17:15 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:15 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:15 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:15 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:15 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:15 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:16 center kernel: ppp: write frame, count = 88 Jan 27 23:17:16 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:16 center kernel: 00 B9 00 00 40 01 15 0C ....@... Jan 27 23:17:16 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:16 center kernel: 08 00 B5 49 04 02 05 00 ...I.... Jan 27 23:17:16 center kernel: FC 6A CE 34 76 11 0E 00 .j.4v... Jan 27 23:17:16 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:16 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:16 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:16 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:16 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:16 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:17 center kernel: ppp: write frame, count = 88 Jan 27 23:17:17 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:17 center kernel: 00 BA 00 00 40 01 15 0B ....@... Jan 27 23:17:17 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:17 center kernel: 08 00 31 51 04 02 06 00 ..1Q.... Jan 27 23:17:17 center kernel: FD 6A CE 34 F8 09 0E 00 .j.4.... Jan 27 23:17:17 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:17 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:17 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:17 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:17 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:17 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:18 center kernel: ppp: write frame, count = 88 Jan 27 23:17:18 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:18 center kernel: 00 BB 00 00 40 01 15 0A ....@... Jan 27 23:17:18 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:18 center kernel: 08 00 B0 49 04 02 07 00 ...I.... Jan 27 23:17:18 center kernel: FE 6A CE 34 77 11 0E 00 .j.4w... Jan 27 23:17:18 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:18 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:18 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:18 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:18 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:18 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:17:19 center kernel: ppp: write frame, count = 88 Jan 27 23:17:19 center kernel: FF 03 00 21 45 00 00 54 ...!E..T Jan 27 23:17:19 center kernel: 00 BC 00 00 40 01 15 09 ....@... Jan 27 23:17:19 center kernel: CC F4 65 01 CC F4 65 FA ..e...e. Jan 27 23:17:19 center kernel: 08 00 32 51 04 02 08 00 ..2Q.... Jan 27 23:17:19 center kernel: FF 6A CE 34 F3 09 0E 00 .j.4.... Jan 27 23:17:19 center kernel: 08 09 0A 0B 0C 0D 0E 0F ........ Jan 27 23:17:19 center kernel: 10 11 12 13 14 15 16 17 ........ Jan 27 23:17:19 center kernel: 18 19 1A 1B 1C 1D 1E 1F ........ Jan 27 23:17:19 center kernel: 20 21 22 23 24 25 26 27 !"#$%&' Jan 27 23:17:19 center kernel: 28 29 2A 2B 2C 2D 2E 2F ()*+,-./ Jan 27 23:17:19 center kernel: 30 31 32 33 34 35 36 37 01234567 Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set flags to f07004b Jan 27 23:18:08 center kernel: ppp: channel ppp0 going down for IP packets! Jan 27 23:18:08 center kernel: ppp0: ccp closed Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set flags to f07000b Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set xmit asyncmap ffffffff Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set flags to f070008 Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set mru to 5dc Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set rcv asyncmap ffffffff Jan 27 23:18:08 center kernel: ppp_tty_ioctl: set flags to f070008 Jan 27 23:18:08 center kernel: ppp: write frame, count = 20 Jan 27 23:18:08 center kernel: FF 03 C0 21 05 02 00 10 ...!.... Jan 27 23:18:08 center kernel: 55 73 65 72 20 72 65 71 User req Jan 27 23:18:08 center kernel: 75 65 73 74 uest Jan 27 23:18:08 center pppd[500]: Terminating on signal 15. Jan 27 23:18:08 center pppd[500]: sent [LCP TermReq id=0x2 "User request"] Jan 27 23:18:09 center kernel: ppp: channel ppp0 closing. Jan 27 23:18:09 center kernel: ppp0 released Jan 27 23:18:09 center kernel: ppp0: ccp closed Jan 27 23:18:09 center pppd[500]: Modem hangup Jan 27 23:18:09 center pppd[500]: Connection terminated. Jan 27 23:18:10 center pppd[500]: Exit. After add the following command my kernel table is below! > route add default ppp0 ( I can also use defaultroute instead ) Kernel routing table Destination Gateway Genmask Flags MSS Window Use Iface 204.244.101.250 * 255.255.255.255 UH 1500 0 2 ppp0 204.244.101.0 * 255.255.255.248 U 1500 0 0 eth0 127.0.0.0 * 255.0.0.0 U 3584 0 8 lo default * * U 1500 0 0 ppp0 Thank you! -- Wesley Shih mailto:hshih@smartt.com

[Quoth Tatai SV Krishnan] ] ] A lot of you are interested in OSPF. I would like to start a discussion ] on this subject. Namely I would like to know from you the following ] ] a. What type of features that you would like to see in TC supporting OSPF I would like to see an ARC or Netserver cabable of being at least the BDR on a network with up to 30 or 40 other OSPF-capable routers. I only need my boxes to be be in a single OSPF area. I don't need NSSA, but I would really like to see OSPF-based load-balancing, a la Ascend. ] b. Any specific authentication types MD5 message-digest. ] c. Your comments on how your networks are setup and the products/problems ] that you see in your current installation. I have the problem of thousands of RIP routes, although I can kludge around that with the ripv2 hack I posted here a few days back. Other than that, it's route convergence time that's my big problem. My networks at all my POPs are all total controls talking with a Cisco. In some cases, I have the ethernet interfaces in a small subnet, in others, I have one big network for the ethernet and the dialup pools. I have static IPs that I allow to work from multiple POPs. -- Michael Mittelstadt meek@execpc.com VP - Internet Techologies ExecPC Internet http://www.execpc.com/~meek 1-800-ExecPC-1

Thus spake Tatai SV Krishnan >On Thu, 29 Jan 1998, Jeff Binkley wrote: >NETServer supports MPIP, The NETServer doing MPIP should have the same >version of code. One NETServer should be the master and client the other >netservers should be the client. The secret should match and also the >time server should be set on all NETServer. If this is not done then you >will have problems only on MPIP calls not on all ISDN calls. And even if this is all done, you still could have problems with all MLPPP calls (at least we are...call reference 27457, open since Dec. 31). I'm quickly coming to the conclusion that MPIP just plain doesn't work reliably. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

I have a couple of users who have Sportster 840s that were chip upgraded from 33.6 to x2. Neither of them can connect to a USR TC rack at x2 speeds. They just get V.34 connects. One of them can connect to a USR MP/8I at x2 speeds just fine. The other one we haven't tried with the MP/8I yet. I noticed some people with the same modem model complaining that they don't get x2 connects in the comp.dcom.modems newsgroup. Are there any known problems with the chip upgraded Sportster 840 connecting at x2 speed? Anyone else seeing this problem? I called USR, and they said to call the Sportster tech support. Brian

Hello all, We are using the TCS 3.0 Total Control Netservers and things have been working well. I have a device, however, that I have not been able to connect with, a CicsoPro 752 ISDN router. I have was reading the Netserver documentation and I came across some text that said something to the effect of "If you are using an ISDN adapter, such as a router but not a modem, you will have to connect using V.120, CHAP, or synchronous PPP becasue the Netserver will not authenticate asynchronously. I know that CHAP is not enabled on our Netservers, so I was wondering if anyone out there has sucessfully set up a CiscoPro ISDN router using one of the other two protocols, or is there yet another option that I missed? Thank you very much, - Richard.

I currently have a few USR MP/16 units which I'm phasing out in favor of USR TC racks. One customer has Windows 3.1 with Trumpet Winsock. When she starts Netscape, Trumpet dials as it should. When she exits Netscape is where the problems start. When she calls our MP/16, Trumpet hangs up just fine after Netscape is exited. When she changes the phone number to call our USR TC rack, she claims to gets no dialtone on any of her phones, even with the computer off. (I think it's a internal modem.) She said she picked her phone an hour after she logged off, and her phone was dead. We have a 15 minute idle timeout, so even if her modem was still connected, it would have hung up long before an hour was up. We told her to switch to Trumpet and hit the bye command before she exits Netscape, but she's bitching and moaning about having to do that extra step. Anyone seen anything like this before? She says if she changes back to the old phone number, that it works fine again. Brian

We installed a HyperARC with release code 4.0.19 on it the other day. Our tech support has been flooded with people calling saying that there windows95 "cannot negotiate a compatible set of protocols". Our tech support has the customers re-load tcp/ip onto there system (like from the win95 cdrom) and alot of the customers are fixed. From what I have been told they are doing this on the order of 20+ users a day, and there is too much. Is there an explaination or fix for this? Also I was quite concerned that users dialing into the Hyperarc come accross radius as authenticated "dumb" connections, any ideas (they are using pap authentication) Perhaps the latest ER fixes this problem? is this even a known occurence with HyperARC or no? 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 -----------------------------/

On Thu, 29 Jan 1998, Jeff Binkley wrote: > > > > U>As you know we installed a hiperarc yesterday with few problems. At > U>first, we forgot to turn off MPIP which screwed up ISDN PPP > U>negotiations. And a few subnets are problematic for which we have an > U>update that fixes framed-routes. > U>However we are experiencing a new problem that suddenly affects about > U>25% of our users.. And we can fix it about 75% of the time.. It > U>corresponded to the HiperARC upgrade but affects all chassis.. > > Can someone give a 2 minute explanation on why MPIP affects ISDN PPP > connections and if that also holds true to Netservers ? it doesnt, it effected all MPP calls. Our first hub was set as the MPIP server, and the other hubs were "clients". The first hub was switched to a HyperARC, and so any MPP calls coming into the other hubs, they would look for the MPIP server to see about PPP bundles, and the call would fail since the MPIP server was not up and operational. If you declare an MPIP server, and that server is not up, your MPP calls will fail. Brian > > Jeff Binkley > ASA Network Computing > > CMPQwk 1.42 9999 > > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

Krish, We're still having problems getting the DOV calls to come in right. Last weekend you said you would have a sample setup available to you there -- I may have missed it. Thnaks for your help, DCP

Brian said once upon a time: > >We installed a HyperARC with release code 4.0.19 on it the other day. Our >tech support has been flooded with people calling saying that there >windows95 "cannot negotiate a compatible set of protocols". Our tech >support has the customers re-load tcp/ip onto there system (like from the >win95 cdrom) and alot of the customers are fixed. We haven't been seeing that here Brian, although I would like to know the exact cause of that error. I have seen that before.

At 05:27 AM 1/28/98 -0600, Tatai SV Krishnan wrote: >On Wed, 28 Jan 1998, Pete Ashdown wrote: > >> Tatai SV Krishnan said once upon a time: >> >> >Each HiPer ARC with the current code supports 7 HDM - you can have two >> >HiPer ARC in one chassis, but make sure that you disable chassis awarness >> >on the HiPer ARCS. "disable nmc chas " >> >Also you must configure each HiPer ARC to be owner of only 7 HDM. >> >> Why am I running 10? Or does this apply to chassis with two ARCs? > >The HiPer ARC is capable of having 10 HDM in a chassis. The current >release of code - we suggest users to upto 7 modems. Also you can have 2 >HiPer ARC in one chassis. > This is different from the original specs I heard on the HiPer ARC. From a message early in the beta testing: The HiPer Access Router Card is a RISC based NAC that can support up to 14 HiPer DSP NACs in one chassis. It is our next generation Router card for supporting high call density and high call volume. It can also be used with the older Quad modem cards. The recommended maximum configuration will have 14 HiPer DSPs, and 2 HiPer ARCs, which will allow for redundancy or load sharing. I believe your comments on this list are the first I've seen, from an official USR/3-Com source that are contrary to the above...but maybe I missed something. 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...

Do any of you support LAT protocol in total control enterprise hubs? If so, can the netserver gateway (or encapsulate) to TCP/IP and what is necessary to set this up? ========================================================================= 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

Richard Stanaford said once upon a time: > We are using the TCS 3.0 Total Control Netservers and things have >been working well. I have a device, however, that I have not been able to >connect with, a CicsoPro 752 ISDN router. I have was reading the >Netserver documentation and I came across some text that said something to >the effect of "If you are using an ISDN adapter, such as a router but not >a modem, you will have to connect using V.120, CHAP, or synchronous PPP >becasue the Netserver will not authenticate asynchronously. I know that >CHAP is not enabled on our Netservers, so I was wondering if anyone out >there has sucessfully set up a CiscoPro ISDN router using one of the other >two protocols, or is there yet another option that I missed? This bug was addressed by Mike Wronski earlier this month. I suggest you check the archives at ftp.xmission.com:/pub/lists/usr-tc.

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?

Brian said once upon a time: > >I am using the release version of the HyperARC code. I am also using the >latest version of Merit RADIUS. > >All of the users who are authenticated and log on the chassis with the >Hyperarc in it show "dumb" as there login type, is this right? Depends on what they're doing. Mine show "rlogin".

Allen Marsalis said once upon a time: >PROBLEM: >Some users can't connect.. They receive the following error.. >This happens on both arc equipped and netserver equipped chassis.. I would go through *ALL* the settings on your HDM's and verify they are set to the defaults, except for the ISDN settings I posted to the list the other day. I had some HDMs shipped to me configured to connect at only 300 BPS!! The problem drove me crazy until I went through all the settings and verified they were correct. The HDM template makes this fairly easy.

At 08:40 PM 1/28/98 -0600, Tatai SV Krishnan wrote: > >A lot of you are interested in OSPF. I would like to start a discussion >on this subject. Namely I would like to know from you the following > >a. What type of features that you would like to see in TC supporting OSPF >b. Any specific authentication types >c. Your comments on how your networks are setup and the products/problems > that you see in your current installation. > a. I would like to see OSPF w/ smart VLSM support. That is, for a netserver to have the ability to take a pool of addresses assigned to it and split it up into 5-6 VLSM nets and advertise those to cut the size of routing tables. Of course NSSA support and different Areas support. OSPF on Async and Sync ports would be nice, but is not urgent. b. MD5 would be fine. As long as it works with CISCO's authentication. c. The biggest problem by far right now is the fact that since there is no OSPF suppot, having multiple TCs puts a huge strain on my Cisco 2501 as far as RIP is concerned. The routing tables are huge and very hard for 2501 to deal with considering they change frequently. If you guys can implement OSPF in the near future, that would give you one more advantage over PM3s. That, quake lag fix and in-netserver DOV support will convince me to dump Livingston and stay with TCs across the border. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 08:34 AM 1/29/98 -0800, you wrote: >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-encapsulated 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 > You have to incorporate our dictionary into Merit. I don't have the time to give you the exact details.. But it can be done.. We send a lot of information in VSA's.. And "NO" you can't control what information we send out of the NS. The format of the packet is as described in the RADIUS RFC.. the first 32 bits are vendor type stuff (vUSR-0000) Then comes the VSA attribute number and its value. The attribute number is always 2 octets and the value is always 4 octets. For example : vUSR-000090230000000c has an attribute number of 0x9023 which translates to Initial-Connect-Rate with a value of 0x0c or 19.2k (I think). Hope this helps. -m

I've seen this problem over the last year or so occasionally with our users. First let me describe our implementation of the tc in our network. (laughter is an ok response) Due to funds limitations (this was before the specials) we couldn't afford the full kit, so what I ended up with is the tc chassis, dual 45 Amp power supplies, 1 dual T-1 card, 6 quad digital/analog modems, 1 nmc with a token ring interface (too bad I've ended up with a ethernet network), no management software, and 2 digi ccon 16s feeding into an NT server running Win NT 3.51 RAS. (We've disabled the T-1s last January and currently the modem are being fed by an analog centrex line pool) Line quality out here in the redwood forest being so lousy as it is, the digital trunks were not worth the cost, 24,000 bps is the average home user connect speed. So yes this is a nightmare kludge, but it actually works. Ok, so you can assume, when the problems come up there is no automatic recovery, fortunately the lease expires in a year. (For a look at the shop http://www.wclynx.com/links/lair.htm quickcam with a 30 second update) Ok... to the problem. For the basic ppp connection, three things are required in the network control panel. 1. Client for MS networks (saves password and other configurations) 2. Dial-Up Adapter 3. TCP/IP protocol (for internet, add/substitute whatever protocol for your lan) Corruption in any of these three components will generate the cannot negotiate compatible a compatible set of protocols error. And since these are loaded into working memory, somtimes all that is required is a cold reboot. (modem needs a reset as well) So the question is where can the corruption come from. First suspect, improper system shutdown. Second power line glitches, even home users need UPSs these days. Third, recent browser upgrades. MS has released a new version of it's winsock implementation, and it affect the dial-up networking. When you reload the TCP/IP protocol, you are replacing the modified stuff with the original distribution version. This is rare, that the browser upgrade damages the networking files, but it does happen. Power hits are still number one out here. With the case of the different chassis on the ISP end behaving differently, I would suspect that the implementation of the HyperARC's winsock is just different enough from the new MS code to cause problems. If you wish I'll do some research into the behavior differences of the two on the MS end. > From: Pete Ashdown <pashdown@xmission.com> > Subject: Re: (usr-tc) HyperARC/Protocol Troubles > To: usr-tc@lists.xmission.com > Date: Thu, 29 Jan 1998 10:48:15 -0700 (MST) > Reply-to: usr-tc@lists.xmission.com > Brian said once upon a time: > > > >We installed a HyperARC with release code 4.0.19 on it the other day. Our > >tech support has been flooded with people calling saying that there > >windows95 "cannot negotiate a compatible set of protocols". Our tech > >support has the customers re-load tcp/ip onto there system (like from the > >win95 cdrom) and alot of the customers are fixed. > > We haven't been seeing that here Brian, although I would like to know the > exact cause of that error. I have seen that before. > > - Dave Sherry aka... Malevil@wclynx.com Webmaster@wclynx.com Postmaster@wclynx.com support@ap.net (707) 887-4640 (voice) (707) 887-1810 (fax) "Computers don't just die, They're murdered!"

John Campbell said once upon a time: > > We just upgraded our system from 1 PRI to 2 PRI's. When we did this, our >ISDN connections went to busy signals even though there are plenty of open >ports. Does anyone have the configuration for the Total Control Hub to >verify that all the ISDN is working correctly. Any help or advice would be >appreaciated. Run the performance monitor on your PRIs and see what channels are in use. ISDN channels will have slot 16 as their gateway, if you are using the Netserver for ISDN.

With the CISCO 766 routers we had to add %isdn to the username (joeblow%isdn) on the router in order for the TC to terminate the call. Also I have read that you need to terminate on the Munich card and not the quadmodems with this router and a few others. If you like to see our UPL file for the router I will e-mail it to you. Good Luck Mike Hamrich DrFast.Net, Inc.

We are sending our acc log to Radius 2.1 on LINUX rather than to the NT versions that ships with the TC. It seems that the logs are missing a few events. What would cause this and are there any settings to be made that are not obvious. Mike Hamrich DrFast.Net Inc.

Livingston recommends for the PM3 with Stac card that modem users not run V.42Bis in conjunction with Stac compression. Is the same recommendation true for the Netserver? If I want to want to turn off just Stac compression, what option do I use with the 'set ccp' command. Is there any compression besides Stac worth keeping? I'm not using ISDN with my USR gear, and I'm not sure that Stac is of any real advantage for modem users. Brian

At 11:21 AM 1/29/98 -0600, Mike Wronski wrote: >You have to incorporate our dictionary into Merit. I don't have the time to >give you the >exact details.. But it can be done.. We send a lot of information in >VSA's.. And "NO" you can't control what information we send out of the NS. > >The format of the packet is as described in the RADIUS RFC.. >the first 32 bits are vendor type stuff (vUSR-0000) Then comes the VSA >attribute number and >its value. The attribute number is always 2 octets and the value is always >4 octets. >For example : >vUSR-000090230000000c has an attribute number of 0x9023 which translates to >Initial-Connect-Rate with a value of 0x0c or 19.2k (I think). > Can you please give us the entries that we need to put into vendors file of Merit Radius for it to recognize those things. Because, I personally do not know what exactly is suppose to go there.. Thanks Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Thu, 29 Jan 1998, Garry Shtern wrote: > c. The biggest problem by far right now is the fact that since there is no > OSPF suppot, having multiple TCs puts a huge strain on my Cisco 2501 as far > as RIP is concerned. The routing tables are huge and very hard for 2501 to > deal with considering they change frequently. If you guys can implement > OSPF in the near future, that would give you one more advantage over PM3s. > That, quake lag fix and in-netserver DOV support will convince me to dump > Livingston and stay with TCs across the border. I'm curious. For what reasons would you dump Livingston and go with with all USR, if the features were basically equal? I have several USR TCs and a PM3. The PM3 is a few days old, so I can't say much about that yet. Brian

> A lot of you are interested in OSPF. I would like to start a discussion > on this subject. Namely I would like to know from you the following > > a. What type of features that you would like to see in TC supporting OSPF > b. Any specific authentication types > c. Your comments on how your networks are setup and the products/problems > that you see in your current installation. Krish, From our perspective, I would like a solid, modest featured ospf, on a rather sooner time schedule. This would support basic VLSM, and simple password authentication but not necessarily more extensive, features like auto VLSM aggregation, etc. Increasing complexity will not only slow down the development, but also the testing and cleanup intervals. OSPF with basic VLSM would be a big step forward for us. Time enough later for the ospf feature arms race between vendors. (granted some of these features will indeed be incremental improvements, but should not come at the expense of the basic widely needed capability) Ken Leland Monmouth Internet

At 01:18 PM 1/29/98 -0600, Brian Elfert wrote: > > >On Thu, 29 Jan 1998, Garry Shtern wrote: > >> c. The biggest problem by far right now is the fact that since there is no >> OSPF suppot, having multiple TCs puts a huge strain on my Cisco 2501 as far >> as RIP is concerned. The routing tables are huge and very hard for 2501 to >> deal with considering they change frequently. If you guys can implement >> OSPF in the near future, that would give you one more advantage over PM3s. >> That, quake lag fix and in-netserver DOV support will convince me to dump >> Livingston and stay with TCs across the border. > >I'm curious. For what reasons would you dump Livingston and go with with >all USR, if the features were basically equal? > >I have several USR TCs and a PM3. The PM3 is a few days old, so I can't >say much about that yet. Well, for starters PM3s do not have TCM nor will they have anything even remotely resembling that in the near future even when Betty comes out with ComOS 3.8 which they are not even giving the release date on. Second, Total Controls use the equivalent of courriers in the quad and hyper cards, and there is nothing out there that even comes close. Thirdly, and probably most importantly, we want to have a uniform platform, since there is not going to be KFlex and X2 battle in couple of months anymore. I can not say anything bad about Livingston's equipment -- it does it's job. However, USR is a vendor of choice at this ISP. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Thu, 29 Jan 1998, Garry Shtern wrote: > >I'm curious. For what reasons would you dump Livingston and go with with > >all USR, if the features were basically equal? > > > >I have several USR TCs and a PM3. The PM3 is a few days old, so I can't > >say much about that yet. > > Well, for starters PM3s do not have TCM nor will they have anything even > remotely resembling that in the near future even when Betty comes out with > ComOS 3.8 which they are not even giving the release date on. At Livingston is honest with release dates. When I bought my first TC rack, I was told OSPF by end of Q3 '97. Then, it slipped to year end '97, and now it's summer '98. > Second, Total Controls use the equivalent of courriers in the quad and > hyper cards, and there is nothing out there that even comes close. This is one of the big reasons I've purchased the USRs to date. Modem compatability is nearly perfect. > Thirdly, and probably most importantly, we want to have a uniform platform, > since there is not going to be KFlex and X2 battle in couple of months > anymore. It is nice to have one brand, but I can't stand the lack of OSPF on the USR, and the amount of heat the USR racks generate. The HiperDSP and HiperARC pretty much take care of the heat issue, (They generate far less heat per modem.) but OSPF is still way out there. I think I'm going to buy Livingston exclusively once we have a standard, and just use the USR gear for those have problems connecting to the PM3. Brian

Charles Hill said once upon a time: > > >Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >the Netserver Ethernet NIC? I have my Netservers running with fullduplex to the hub with no problems. I believe full duplex is part of the 100 TX spec, which is what the ARC is. I'm getting a Catalyst 5500 in the next month or so, and this is one of the first things I'm going to try with the ARC.

Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about the Netserver Ethernet NIC? Regards, Charles Hill ioNET Network Specialist work: 405.270.7020 cell: 405.833.5477 pager: 405.559.6697 or 4058335477@mobile.att.net email: chill@ionet.net http://www.ionet.net/~chill ICQ: 4923465@pager.mirabilis.com

At 05:16 PM 1/29/98 -0600, Charles Hill wrote: > >Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >the Netserver Ethernet NIC? > >Regards, > >Charles Hill My understanding is that the netservers 10T port is full duplex capable in hardware but only half duplex is supported by recent netserver code. (Unless they added it in the last several months.) I have found that full duplex 10T is sort of rare. Both the USR TotalSwitch (discontinued?) and the 3COM superstack switch are half duplex on the 10T ports. We purchased a Catalyst just for full duplex 10T. As for 100T, I have never seen a 100T port that wasnt full-duplex capable. (except on a hub of course) The superstacks lone 100T port is full duplex even though the 10T ports are half.. So I would hope that the hiperarcs 100T ports (and code) fully support full duplex.. Should be easy to check.. Enable full duplex at the switch and watch the switch for crc errors, slowness, etc.. 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

At 05:16 PM 1/29/98 -0600, Charles Hill wrote: > >Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >the Netserver Ethernet NIC? > >Regards, > >Charles Hill My understanding is that the netservers 10T port is full duplex capable in hardware but only half duplex is supported by recent netserver code. (Unless they added it in the last several months.) I have found that full duplex 10T is sort of rare. Both the USR TotalSwitch (discontinued?) and the 3COM superstack switch are half duplex on the 10T ports. We purchased a Catalyst just for full duplex 10T. As for 100T, I have never seen a 100T port that wasnt full-duplex capable. (except on a hub of course) The superstacks lone 100T port is full duplex even though the 10T ports are half.. So I would hope that the hiperarcs 100T ports (and code) fully support full duplex.. Should be easy to check.. Enable full duplex at the switch and watch the switch for crc errors, slowness, etc.. 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

At 08:40 PM 1/28/98 -0600, you wrote: > >A lot of you are interested in OSPF. I would like to start a discussion >on this subject. Namely I would like to know from you the following > >a. What type of features that you would like to see in TC supporting OSPF >b. Any specific authentication types >c. Your comments on how your networks are setup and the products/problems > that you see in your current installation. Any implementation of OSPF would be extremely welcome. I'd hold off on fancy features, and get out a solid release as quickly as possible. Our greatest need is to get rid of the ridiculous RIP routes that put such a load on our routers. At each POP, we have a Cisco 2501 along with a combination of access servers: Cisco 2511's, PM2e's, and USR TCs. EVERYTHING can nicely run OSPF except the TCs. This and the UDP latency problems are by far my biggest gripes with the TCs. I wish I could combine the TC's modems with Cisco's IOS features and a PM3's ease of configuration. THAT would be a nice solution! 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 _____________________________________________

On Fri, 30 Jan 1998, Martin Pfeilsticker wrote: > Hello ! > > I have a problem with a USR Netserver 16/I and callback. > > We have a customer, who wants to dial in with one B-Channel ( X.75) > and then is called back with two B-Channels. > > I created a location and a simple Dial-Script ( ATD1234, expect > "CONNECT"). In the location "High Mark" is set to 10, "Idle > Time-out" is the default, 0. > In the dialgroup are the ports 5 and 6, I have tried other ports as > well. Both ports are working without any problems. Well first of all this will not supported since the NETServer does not do multilink dialout. However you can make it work with some changes. First your location should have atleast 2 max ports and they should belong to the same group. So say you set s5 and s6 as login network twoway, then you have to set s5 and s6 to one group - say 1. You location should also have the same so set location < name > group 1 maxports 1 and set the the high water mark to 1. When you set it to 10 you will have to wait till you reach that threshold. Now take a look at your script, you are asking one B channel to dial 1234, the first channel is going to dial that and connect - what will happen if the second channel tries to dial the same number? Atleast here in the US it will give you a busy signal. It also depends upon what type of ISDN line you have and your TA to see if you do need two numbers. If busy is not a problem and that you are sure that you can dial the same number and connect your script does not need any modification. Else you need to store the number on your S5 and S6 port and dial the stored numbers. Now when the first link is connected you can force the second link to connect also by either manually logging on the the NETServer and asking it to dial the location or just by setting the high water mark. Typically this will connect and make a 2 B channel connection, but sometimes based on the type of negotiation you may have start one channel first and the second channel next. This is the way to do it, and this is not supported. krish > > When the customer calls in, he is authenticated and receives > the callback, but only from one channel. The second channel > is never opened, and even worse, it doesn't even tries to dial out. > > I have tested both ports with callback, and regardless of the > portnumber, the behavior is same, only one channel is opened. > > Does anybody have any suggestions ? I have tried to reach the > european hotline of USR/3Com, but they seems to have a different > concept of "We will call you back tomorrow" as the rest of the world. > > Thanks > Martin Pfeilsticker > > > > > > > Martin Pfeilsticker Product Support Engineer > Pan Dacom Daten- und Kommunikationssysteme GmbH > Robert-Bosch-Strasse 32 63303 Dreieich, Germany > Tel.: x49-6103-932-156 > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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

Once upon a time Garry Shtern shaped the electrons to say... >Well, for starters PM3s do not have TCM nor will they have anything even >remotely resembling that in the near future even when Betty comes out with >ComOS 3.8 which they are not even giving the release date on. Remember when you chimed in with incorrect statements on RADIUS ABM, aka Betty, on our list and got shut down. You have never seen the current version of Betty, or you are violating an NDA by talking about it. QED. Also, Betty is in no what a chassis management utility - that is what 'Amber' is. The next generation PMConsole - 100% Java, with a slew of new features. And the same thing that goes for Betty goes for Amber, so how can you comment on it? And no, we don't give release dates for software in development. Never have. Personally I don't care what you buy, but I do care when FUD is spread. You cannot intelligently comment on products that you have not seen, and if you had seen them you still couldn't comment without violating either an NDA or beta agreement. There were early versions of Betty and Amber shown at some tradeshows which gave people some idea of where it is going, but they are nearing release now and have expanded a great deal beyond anything the public has seen. -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

Try setting dip switch 5 to on and reinserting the NMC card this will reset it factory defaults.

Brian Elfert wrote: > > I currently have a few USR MP/16 units which I'm phasing out in favor of > USR TC racks. > > One customer has Windows 3.1 with Trumpet Winsock. When she starts > Netscape, Trumpet dials as it should. When she exits Netscape is where > the problems start. How about having her download the Netscape 3.x with built in dialer? May not address the problem but, if it solves it maybe you will not have to deal with her. We get a lot of these types of people that can't take an extra mouse click or two. Like when Win95 forgets to load the dun dialer for netscape but, not MSIE. A simple registery hack is all it is but, I am not taking customers in to the registry as if they screw something up... Anyway it's just a though.

I just flashed our hiperarc with new code krish sent me. (ne040086.dmf) Thanks krish.. I also reflashed our hdm's because this is where brian feels our ppp negotiation problems lie.. The download would fail when I tried all four hdms at once, so I flashed them one at a time. Other than that, it went real smooth. Finding the hdm code was a real pain. The totalservice site was down with the message; "Can't connect to local MSQL server".. I could locate the file "hd010007.zip" via ftp but couldn't download it, with or without passwords. I couldn't find it on the cd either. I phoned usr and the tech said he thought it was on reverb.. So I grabbed the file off of reverb.. However the hyperlink said: T1 HDM Modem 1.0.7 10-14-97 Beta #3 code which makes me sort of nervous.. Does anyone know if this is the release code under the same filename on totalservice?.. Perhaps beta 3 was the final release version.. Just want to confirm that I used the right code. If it is not the right file, will someone please email me the current file. (or fix sql..) Also I wanted to confirm that this file flashes both the T1 code and the modem code on the hdm.. I saved both to nvram before resetting. But I wasn't sure.. Thanks for the help.. 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

Looking for a list, of addresses and Phone Numbers for Resellers of USR Total Control Chassis! Best deal I have found so far is 12,000 on the 2509 Chassis! Anyone found a better deal? You may reply privately to jault@globalco.net tech@globalco.net jault@globalco.net GlobalNet Member ISP/C

Has anyone seen an issue with dialup service logging into a shell host, where exiting from pico and saving changes will cause the connection to drop. I never got complaints when using past netserver code like 3.2.27 etc. only when using tcs251 and higher. This happens about 1 out of 10 editing attempts. any ideas???????? ______________________________________________________________________ | Henry Moats Network Services Support nc0419 ext 3671 | ______________________________________________________________________|

Hello ! I have a problem with a USR Netserver 16/I and callback. We have a customer, who wants to dial in with one B-Channel ( X.75) and then is called back with two B-Channels. I created a location and a simple Dial-Script ( ATD1234, expect "CONNECT"). In the location "High Mark" is set to 10, "Idle Time-out" is the default, 0. In the dialgroup are the ports 5 and 6, I have tried other ports as well. Both ports are working without any problems. When the customer calls in, he is authenticated and receives the callback, but only from one channel. The second channel is never opened, and even worse, it doesn't even tries to dial out. I have tested both ports with callback, and regardless of the portnumber, the behavior is same, only one channel is opened. Does anybody have any suggestions ? I have tried to reach the european hotline of USR/3Com, but they seems to have a different concept of "We will call you back tomorrow" as the rest of the world. Thanks Martin Pfeilsticker Martin Pfeilsticker Product Support Engineer Pan Dacom Daten- und Kommunikationssysteme GmbH Robert-Bosch-Strasse 32 63303 Dreieich, Germany Tel.: x49-6103-932-156

At 05:09 AM 1/30/98 -0600, Allen Marsalis wrote: >I could locate the file "hd010007.zip" via ftp but couldn't download it, >with or without passwords. I will let the Web group know about this... >I couldn't find it on the cd either. > >I phoned usr and the tech said he thought it was on reverb.. > >So I grabbed the file off of reverb.. However the hyperlink said: > T1 HDM Modem 1.0.7 10-14-97 Beta #3 code >which makes me sort of nervous.. Does anyone know if this is the release >code under the same filename on totalservice?.. Perhaps beta 3 was the final >release version.. Just want to confirm that I used the right code. If >it is not the right file, will someone please email me the current file. >(or fix sql..) > >Also I wanted to confirm that this file flashes both the T1 code and the >modem code on the hdm.. I saved both to nvram before resetting. But I >wasn't sure.. > Every BETA code is potentially the released code.. We don't use different numbers for BETA & released.. So even though the code says beta of reverb, if it has the same code revision number as was is available on total service it will be the same as the released code.. 1.0.7 is 1.0.7BETA.. -m

I'm a little concerned about the memory leak issue with 3.6.28 that i heard about for 8meg netsv cards. Also, what about standard ethernet nics compared to high speed? It was recommended by my USR tech person that 3.6.28 combined with standard ethernet nic could cause lock ups and reboots for the netsv. I have yet to try this combination in a lab env. has anyone attempted this. if so, any feedback??? thanks On Fri, 30 Jan 1998, Pete Ashdown wrote: > Henry Moats said once upon a time: > > > > > >Has anyone seen an issue with dialup service logging into a > >shell host, where exiting from pico and saving changes will cause the > >connection to drop. I never got complaints when using past netserver > >code like 3.2.27 etc. only when using tcs251 and higher. > >This happens about 1 out of 10 editing attempts. > > Upgrade to 3.6.28. There is a buffer IO problem with 3.4 -> 3.5 that > causes this type of behavior with some programs like Pico and rzsz. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the message. > For information on digests or retrieving files and old messages send > "help" to the same address. Do not use quotes in your message. >

> So I grabbed the file off of reverb.. However the hyperlink said: > T1 HDM Modem 1.0.7 10-14-97 Beta #3 code > which makes me sort of nervous.. Does anyone know if this is the release > code under the same filename on totalservice?.. Perhaps beta 3 was the final > release version.. Just want to confirm that I used the right code. If > it is not the right file, will someone please email me the current file. > (or fix sql..) If I remember correctly, there were 3 versions of beta code. They started at 1.0.3, then 1.0.5, and 1.0.7 was the final beta before it was released. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

Hello, I trying to setup the mpipserver on a usr tc and am receiving the following message on my syslog server: Jan 30 10:04:10 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:12 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:14 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:16 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:18 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:20 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:22 usrtc1 [MPIP_SERVER_PING_REQ] Jan 30 10:04:24 usrtc1 [MPIP_SERVER_PING_REQ] Does anyone know what this means? thanks, mark Lehigh University

Henry Moats said once upon a time: > > >Has anyone seen an issue with dialup service logging into a >shell host, where exiting from pico and saving changes will cause the >connection to drop. I never got complaints when using past netserver >code like 3.2.27 etc. only when using tcs251 and higher. >This happens about 1 out of 10 editing attempts. Upgrade to 3.6.28. There is a buffer IO problem with 3.4 -> 3.5 that causes this type of behavior with some programs like Pico and rzsz.

I still have strange reboots periodically with 3.6.28 and 16 meg. It isn't a big enough problem for me to try and wade through support on it though. I hope that soon I'll have banished my Netservers to any one of the rings of hell that they are suited for. The ARC is really a much better product. Henry Moats said once upon a time: > > >I'm a little concerned about the memory leak issue with 3.6.28 >that i heard about for 8meg netsv cards. >Also, what about standard ethernet nics compared to high speed? >It was recommended by my USR tech person that 3.6.28 combined with >standard ethernet nic could cause lock ups and reboots for the >netsv. I have yet to try this combination in a lab env. has anyone >attempted this. if so, any feedback??? > >thanks > > >On Fri, 30 Jan 1998, Pete Ashdown wrote: > >> Henry Moats said once upon a time: >> > >> > >> >Has anyone seen an issue with dialup service logging into a >> >shell host, where exiting from pico and saving changes will cause the >> >connection to drop. I never got complaints when using past netserver >> >code like 3.2.27 etc. only when using tcs251 and higher. >> >This happens about 1 out of 10 editing attempts. >> >> Upgrade to 3.6.28. There is a buffer IO problem with 3.4 -> 3.5 that >> causes this type of behavior with some programs like Pico and rzsz. >> >> - >> To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" >> with "unsubscribe usr-tc" in the body of the 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. >

On Fri, 30 Jan 1998, Allen Marsalis wrote: > Well just when all was going well. One hdm is empty while we are > having busys.. OK so I do a "display atstat" and the Modem Connect > column shows N/A for all channels. Doing a reboot or a sinserv > didn't help. I don't want to reboot the hiperarc if I don't have > to. Anyone got a quick answer on a Friday night??.. > > allen On the Hiper arc do a list interface, it will show if the modems are owned by the Hiper arc or not. If its not owned by hiper then you have to change it by using the command set cha sl <slot#> ow y 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

On Thu, 29 Jan 1998, Pete Ashdown wrote: > Charles Hill said once upon a time: > > > > > >Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about > >the Netserver Ethernet NIC? > > I have my Netservers running with fullduplex to the hub with no problems. > I believe full duplex is part of the 100 TX spec, which is what the ARC > is. I'm getting a Catalyst 5500 in the next month or so, and this is one > of the first things I'm going to try with the ARC. But where do you set full duplex at? I don't know of any nic's that auto detect full duplex, even catalyst switches dont autodetect, so surly there must be a way to turn it on. > > - > To unsubscribe to usr-tc, send an email to "majordomo@xmission.com" > with "unsubscribe usr-tc" in the body of the 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 -----------------------------/

> Uhm...minor nit...depends on the catalyst. The higher end ones do > auto-detect and negotiate full and half duplex (our 2926 connected to > sun hme cards does it quite nicely, once you get the needed patches on > the suns) we have a catalyst 3200 which is definitly a high end switch, and I dont beleive it will auto detect. I will try some fdx experiments with the arc later on. > -- > 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 | | 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 -----------------------------/

> I'm a little concerned about the memory leak issue with 3.6.28 > that i heard about for 8meg netsv cards. > Also, what about standard ethernet nics compared to high speed? > It was recommended by my USR tech person that 3.6.28 combined with > standard ethernet nic could cause lock ups and reboots for the > netsv. I have yet to try this combination in a lab env. has anyone > attempted this. if so, any feedback??? No idea on the nic, but DO SOMETHING about your 8mb netserver. Geez, it will only cost around $40 to get a 16MB non-parity SIMM for the thing. | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

It seems that after about 15-30 minutes of being up, our pri get out of sync with the hub. All the calls either get busies or nothing picks up. A hardware reset of the card restores the pri. Any thoughts on the solution? James

Garry Shtern said once upon a time: >>>Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >>>the Netserver Ethernet NIC? >> >>I have my Netservers running with fullduplex to the hub with no problems. >>I believe full duplex is part of the 100 TX spec, which is what the ARC >>is. I'm getting a Catalyst 5500 in the next month or so, and this is one >>of the first things I'm going to try with the ARC. > >Your netservers or Hiper ARCs? I've already tried full-duplex with the Netservers to my XLNT 10BT switch. It seems to work OK, but performance doesn't seem to change with confirms the software issue. I'm going to be attempting full-duplex 100BT with the Catalyst 5500 to the ARC.

Thus spake Brian >On Thu, 29 Jan 1998, Pete Ashdown wrote: >> Charles Hill said once upon a time: >> >Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >> >the Netserver Ethernet NIC? >> I have my Netservers running with fullduplex to the hub with no problems. >> I believe full duplex is part of the 100 TX spec, which is what the ARC >> is. I'm getting a Catalyst 5500 in the next month or so, and this is one >> of the first things I'm going to try with the ARC. fdx is not a part of the 100tx spec...we have several nic cards which are hdx only (sun be cards) >But where do you set full duplex at? I don't know of any nic's that auto >detect full duplex, even catalyst switches dont autodetect, so surly there >must be a way to turn it on. Uhm...minor nit...depends on the catalyst. The higher end ones do auto-detect and negotiate full and half duplex (our 2926 connected to sun hme cards does it quite nicely, once you get the needed patches on the suns) -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

On Fri, 30 Jan 1998, jcondon wrote: > Brian Elfert wrote: > > > > I currently have a few USR MP/16 units which I'm phasing out in favor of > > USR TC racks. > > > > One customer has Windows 3.1 with Trumpet Winsock. When she starts > > Netscape, Trumpet dials as it should. When she exits Netscape is where > > the problems start. > > How about having her download the Netscape 3.x with built in dialer? This would probably fix the problem, but the customer sure as heck isn't gonna want to install new software if they can't stand an extra mouse click now. I'd really like to figure why it works with an MP/16, but not with a TC rack. Brian

Thus spake Brian >> Uhm...minor nit...depends on the catalyst. The higher end ones do >> auto-detect and negotiate full and half duplex (our 2926 connected to >> sun hme cards does it quite nicely, once you get the needed patches on >> the suns) >we have a catalyst 3200 which is definitly a high end switch, and I dont >beleive it will auto detect. I will try some fdx experiments with the arc >later on. Hrmm....perhaps its not a high-end, low-end thing then...I thought it was, but I could be wrong. The 2926 does do auto-detect of full-duplex, but it is basically a fixed-configuration of the 5000 chassis, so it shares many of the features of the 5000. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456

On Fri, 30 Jan 1998, Curt Shambeau wrote: > > So I grabbed the file off of reverb.. However the hyperlink said: > > T1 HDM Modem 1.0.7 10-14-97 Beta #3 code > > which makes me sort of nervous.. Does anyone know if this is the release > > code under the same filename on totalservice?.. Perhaps beta 3 was the final > > release version.. Just want to confirm that I used the right code. If > > it is not the right file, will someone please email me the current file. > > (or fix sql..) > > If I remember correctly, there were 3 versions of beta code. > They started at 1.0.3, then 1.0.5, and 1.0.7 was the final beta before it > was released. Why doesn't USR do like some other vendors, and use a different numbering scheme for betas than for released code? The current system confuses out of some people who think they got ripped off because the version # jumped from say 5.1.7 to 5.7.7, and they wonder why they never got any of the releases inbetween. Brian

Garry Shtern said once upon a time: >>I've already tried full-duplex with the Netservers to my XLNT 10BT switch. >>It seems to work OK, but performance doesn't seem to change with confirms >>the software issue. I'm going to be attempting full-duplex 100BT with the >>Catalyst 5500 to the ARC. > >How exactly did you enable netservers to do full duplex? Plugged them into full-duplex ports on the switch and then reboot them. My experience with full/half ethernet is that if the equipment doesn't support it, it will eventually fail and stop passing packets. This hasn't happened with the Netservers.

At 04:32 PM 1/29/98 -0700, Pete Ashdown wrote: >Charles Hill said once upon a time: >> >> >>Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >>the Netserver Ethernet NIC? > >I have my Netservers running with fullduplex to the hub with no problems. >I believe full duplex is part of the 100 TX spec, which is what the ARC >is. I'm getting a Catalyst 5500 in the next month or so, and this is one >of the first things I'm going to try with the ARC. Your netservers or Hiper ARCs? Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 09:30 AM 1/30/98 -0800, Henry Moats wrote: > >I'm a little concerned about the memory leak issue with 3.6.28 >that i heard about for 8meg netsv cards. >Also, what about standard ethernet nics compared to high speed? >It was recommended by my USR tech person that 3.6.28 combined with >standard ethernet nic could cause lock ups and reboots for the >netsv. I have yet to try this combination in a lab env. has anyone >attempted this. if so, any feedback??? > We have a standard nic and 3.6.28 and everything works without any problems. As for memory leaks, I wouldn't know since we have 20megs in each of them :)) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 12:48 PM 1/30/98 -0500, Jeff Mcadams wrote: >Thus spake Brian >>> Uhm...minor nit...depends on the catalyst. The higher end ones do >>> auto-detect and negotiate full and half duplex (our 2926 connected to >>> sun hme cards does it quite nicely, once you get the needed patches on >>> the suns) > >>we have a catalyst 3200 which is definitly a high end switch, and I dont >>beleive it will auto detect. I will try some fdx experiments with the arc >>later on. > >Hrmm....perhaps its not a high-end, low-end thing then...I thought it >was, but I could be wrong. The 2926 does do auto-detect of full-duplex, >but it is basically a fixed-configuration of the 5000 chassis, so it >shares many of the features of the 5000. Even the cheap-ass BayStack 350 does autodetection. And it is not considered high-end. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 12:17 PM 1/30/98 -0700, Pete Ashdown wrote: >Garry Shtern said once upon a time: > >>>>Will the Ethernet interfaces on the HiPer ARCs do full duplex? What about >>>>the Netserver Ethernet NIC? >>> >>>I have my Netservers running with fullduplex to the hub with no problems. >>>I believe full duplex is part of the 100 TX spec, which is what the ARC >>>is. I'm getting a Catalyst 5500 in the next month or so, and this is one >>>of the first things I'm going to try with the ARC. >> >>Your netservers or Hiper ARCs? > >I've already tried full-duplex with the Netservers to my XLNT 10BT switch. >It seems to work OK, but performance doesn't seem to change with confirms >the software issue. I'm going to be attempting full-duplex 100BT with the >Catalyst 5500 to the ARC. How exactly did you enable netservers to do full duplex? Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Fri, 30 Jan 1998, Garry Shtern wrote: > At 09:34 PM 1/29/98 -0800, MegaZone wrote: > >Once upon a time Garry Shtern shaped the electrons to say... > >>Well, for starters PM3s do not have TCM nor will they have anything even [..] > > I understand you are big lover of Portmasters(after all you work for > Livingston). However, I, and I am sure a lot of other folks here, do not > appreciated the fact that you always stick your 5c into every conversation > where you find you can promote livingston products here. THIS IS NOT A > LIVINGSTON mailing list. On portmaster-users you can promote 3.8 and all > PM3s as much as you heart desires, but this is no place for it. And you > personally should know not to do this because if you honestly think that > Livingston products are better than USR, then you don't need to advertise; > people will find out themselves. My friend, he did not bring up the discussion of our products... Every competitor I know of is on every other competitor's mailing lists. I'd like to think that keeps us all honest and awake.. Most of us are on good terms with each other. I see no reason to change that. No one was promoting anything, Megazone was responding to a statement made about *OUR* products. Anyhow, let's be done with this discussion. We have not yet released any of the products under debate--very true and [I believe] agreed. Notice that none of us (Lucent) chimed in with a promo for them either, someone else asked a question and made a statement about them first--Megazone gave his response. Garry, aren't you on the portmaster-users list? --jr "It's just a computer." --- Josh Richards - <jrichard@livingston.com> Beta Engineer Lucent Technologies (Remote Access Business Unit) (previously Livingston Enterprises, Inc.)

At 09:34 PM 1/29/98 -0800, MegaZone wrote: >Once upon a time Garry Shtern shaped the electrons to say... >>Well, for starters PM3s do not have TCM nor will they have anything even >>remotely resembling that in the near future even when Betty comes out with >>ComOS 3.8 which they are not even giving the release date on. > >Remember when you chimed in with incorrect statements on RADIUS ABM, >aka Betty, on our list and got shut down. You have never seen the current >version of Betty, or you are violating an NDA by talking about it. QED. > I never signed any NDA. You guys brought it to the Internet World exposition here in NYC and that is where I got a chance to play with it and see what it looks like. So, you might say, I have seen the "current" version of betty which was about a month ago. >Also, Betty is in no what a chassis management utility - that is what >'Amber' is. The next generation PMConsole - 100% Java, with a slew of >new features. And the same thing that goes for Betty goes for Amber, >so how can you comment on it? I am sorry, I meant Amber not Betty -- my mistake. However, I did see both of them on the Internet World and neither one impressed me. As far as I am concerned, I have a Windows NT workstation here and I do not care whether it is written in Visual Basic or Java as long as it runs. And TCM does just that -- runs. You guys, however, do not have ANY working version of Amber to distribute and you won't until 3.8 comes out and heck knows when that will happen. >And no, we don't give release dates for software in development. Never >have. > >Personally I don't care what you buy, but I do care when FUD is spread. >You cannot intelligently comment on products that you have not seen, and >if you had seen them you still couldn't comment without violating either >an NDA or beta agreement. There were early versions of Betty and Amber >shown at some tradeshows which gave people some idea of where it is >going, but they are nearing release now and have expanded a great deal >beyond anything the public has seen. Maybe so. But the version we have seen on the trade show did not even come close to what TCM is. And I seriously doubt that the release version does. I understand you are big lover of Portmasters(after all you work for Livingston). However, I, and I am sure a lot of other folks here, do not appreciated the fact that you always stick your 5c into every conversation where you find you can promote livingston products here. THIS IS NOT A LIVINGSTON mailing list. On portmaster-users you can promote 3.8 and all PM3s as much as you heart desires, but this is no place for it. And you personally should know not to do this because if you honestly think that Livingston products are better than USR, then you don't need to advertise; people will find out themselves. Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

I'll make this easy to skip...  Once upon a time Garry Shtern shaped the electrons to say... >exposition here in NYC and that is where I got a chance to play with it and >see what it looks like. So, you might say, I have seen the "current" >version of betty which was about a month ago. No, you saw the demo version we brought to the tradeshow. Which is pale compared to the current beta. Tradeshows are just to gain interest and leads. As an unreleased product we're not going to tip our hand to everyone. The people who have seen the up to date product have an NDA. >it is written in Visual Basic or Java as long as it runs. And TCM does >just that -- runs. You guys, however, do not have ANY working version of >Amber to distribute and you won't until 3.8 comes out and heck knows when >that will happen. Actually Amber is working rather well. Amber is just waiting on 3.8, which will be rather sooner than you seem to think. I know you've seen the *partial* list of features that was posted on portmaster-users, since you followed up that thread. And did you really expect us to give beta code to someone who publically states they'd dump Lucent for 3Com on a mailing list? That sounds like a prime beta candidate to me. >Maybe so. But the version we have seen on the trade show did not even come >close to what TCM is. And I seriously doubt that the release version does. Everything in TCM? No. Some of the analysis features are in Betty since it maintains a full database of use on the backend. But we're not trying to reproduce TCM, because we haven't been asked to buy the customer. We've heard clearly that most of our users are not interested in all of the flash that TCM has, but want a functional configuration utility that allows for all of the logging, debugging, backup/restore, etc abilities they need to keep things running. We also have a diverse userbase with Win95/NT, and nearly every flavor of UNIX available. So cross platform commonality and functionality is a major issue. Great that you have WinNT and it works for you. But have you not seen the other users here asking if/when TCM will be ported to Platform X? It is not enough for them. I'm not going to get into a feature-by-feature comparision, especially on an unreleased product, but what in particular is unforgivably missing tin your mind? >I understand you are big lover of Portmasters(after all you work for >Livingston). However, I, and I am sure a lot of other folks here, do not >appreciated the fact that you always stick your 5c into every conversation >where you find you can promote livingston products here. THIS IS NOT A Frankly I don't care. And, as you know, I have 2 weeks left here before I start at GTE/BBN. And I'll be on from my personal domain after that to be sure - and I'll be less restrained than ever. You don't like it? Filter me out - filter, procmail, or whatever you prefer. The technology exists, so use it instead of complaining. >PM3s as much as you heart desires, but this is no place for it. And you >personally should know not to do this because if you honestly think that >Livingston products are better than USR, then you don't need to advertise; >people will find out themselves. This is bullshit, and history is full of evidence to the contrary. Many good products have failed because of lack of knowledge. And half-baked products have succeeded based on marketing and not technical merit. The 'good products sell themselves' myth has killed many over the years. I don't jump in on every thread, frankly if I did that I'd be in most all of them since there is able opportunity to counter things here. I suggest you check the list history and see how many threads I joined in with - after someone else mentioned Livignston/Lucent/PortMasters/etc directly, as you did in this one. Or when someone makes an indirect comment - "I don't think any product can/can't *blah*". I've said it before - anyone is welcome on the Lucent lists. And Kevin Smith from Ascend has long monitored them and jumped in in exactly the same way when Ascend comes up directly or indirectly. And I'm certainly not singling out this list, as I also monitor the Ascend, Bay, Cisco, Telebit, Shiva, and Microcom mailing lists, to name a few. Again - if you don't like it, use the technology and just filter my email out. Simple. I've always laughed at anyone who whines about my posting, because it is their own damn fault that they read it in the first place. -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

Once upon a time Garry Shtern shaped the electrons to say... >And also a 3 page advertisement of the features of ComOS 3.8! After someone (You I think) made a FUDdy comment about 'not believing Lucent could possibly have a better product than 3Com'. And notice that I made a short reply - followed by a warning and a page break so anyone could simply skip the rest. If your mailer doesn't under stand a simple page break - like this one:  Too damn bad. I pointed out the simple fact that USR came to Livingston to get an OS. That speaks volumes alone. And then, to strengthen the argument I posted the break down. You bet your ass I'm going to counter a comment like that. An easy way to avoid it is to not make them in the first place. -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

> I understand you are big lover of Portmasters(after all you work for > Livingston). However, I, and I am sure a lot of other folks here, do not > appreciated the fact that you always stick your 5c into every conversation > where you find you can promote livingston products here. THIS IS NOT A > LIVINGSTON mailing list. On portmaster-users you can promote 3.8 and all > PM3s as much as you heart desires, but this is no place for it. And you > personally should know not to do this because if you honestly think that > Livingston products are better than USR, then you don't need to advertise; > people will find out themselves. AMEN! Sing it loud! I've found this list to be so valuable to me, but I grow weary of the bickering that goes on from time to time about this product or that. Can we all just talk about USR/3COM products please? | Curtis V. Shambeau | curt@execpc.com | http://www.execpc.com/~curt | | Executive Vice President - Exec-PC, Inc. |

Lee Kuo said once upon a time: >The only times when I have seen pro-Livingston posts from Megazone is >correcting people who have posted incorrect information about Livingston >products. Stick around (or check the archives).

Help please.. I have a customer that has several TC racks exhibiting the same problem. I dial into one of the racks and am assigned an IP address from the pool (203.45.2.x). The netserver that I'm connecting to in the rack is on subnet (192.168.X.X). Routing is enabled on everything and the netserver has valid routes pointing to the local 192.168 network via Net0 and a host based route to my client on 203.45.2.x /32. The problem is I can ping anywhere on the plant EXECPT for any device off the local 192.168.X.X net0 segment (except for the actual Net0 address 192.168.101.31, I can ping that). If however I telnet from my client onto the Netserver and ping any device on the 192.168 subnet, ping responds fine... I have routing on I have enh_routing on I have local_routing on But nothing seems to work... Why can I not ping any other device on the Net0 segment except for net0 itself.. When all routes in the route table say otherwise..?? Please help.. Andrew.

On Fri, 30 Jan 1998, Garry Shtern wrote: > I understand you are big lover of Portmasters(after all you work for > Livingston). However, I, and I am sure a lot of other folks here, do not > appreciated the fact that you always stick your 5c into every conversation > where you find you can promote livingston products here. THIS IS NOT A > LIVINGSTON mailing list. On portmaster-users you can promote 3.8 and all > PM3s as much as you heart desires, but this is no place for it. And you > personally should know not to do this because if you honestly think that > Livingston products are better than USR, then you don't need to advertise; > people will find out themselves. The only times when I have seen pro-Livingston posts from Megazone is correcting people who have posted incorrect information about Livingston products.

on a related topic...does anyone have a simple trumpet setup that works with tc pap logins? matthew At 12:27 PM 1/30/98 -0600, you wrote: > > >On Fri, 30 Jan 1998, jcondon wrote: > >> Brian Elfert wrote: >> > >> > I currently have a few USR MP/16 units which I'm phasing out in favor of >> > USR TC racks. >> > >> > One customer has Windows 3.1 with Trumpet Winsock. When she starts >> > Netscape, Trumpet dials as it should. When she exits Netscape is where >> > the problems start. >> >> How about having her download the Netscape 3.x with built in dialer? > >This would probably fix the problem, but the customer sure as heck isn't >gonna want to install new software if they can't stand an extra mouse >click now. > >I'd really like to figure why it works with an MP/16, but not with a TC >rack. > >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. >

At 04:15 PM 1/30/98 -0600, Lee Kuo wrote: >On Fri, 30 Jan 1998, Garry Shtern wrote: >> I understand you are big lover of Portmasters(after all you work for >> Livingston). However, I, and I am sure a lot of other folks here, do not >> appreciated the fact that you always stick your 5c into every conversation >> where you find you can promote livingston products here. THIS IS NOT A >> LIVINGSTON mailing list. On portmaster-users you can promote 3.8 and all >> PM3s as much as you heart desires, but this is no place for it. And you >> personally should know not to do this because if you honestly think that >> Livingston products are better than USR, then you don't need to advertise; >> people will find out themselves. > >The only times when I have seen pro-Livingston posts from Megazone is >correcting people who have posted incorrect information about Livingston >products. > And also a 3 page advertisement of the features of ComOS 3.8! Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

On Fri, 30 Jan 1998 10:09:30 -0500, Mark Miller <Mark.Miller@Lehigh.EDU> wrote: >Jan 30 10:04:22 usrtc1 [MPIP_SERVER_PING_REQ] >Jan 30 10:04:24 usrtc1 [MPIP_SERVER_PING_REQ] > > >Does anyone know what this means? If I remember right, these are normal, non-critical log messages that fill up your log file every 2 seconds. As for what they mean, I think it's something along the lines of: "usrtc1's MPIP server is sending out a ping request to it's clients" Russ

On Fri, 30 Jan 1998 19:07:27 -0400, James Wilson <james@cruxnet.com> wrote: >Anyone know where I can get the latest release. The website is not ftp'ing >correctly and I need this STAT. > >James >james@cruxnet.com ftp://username:password@totalservice.usr.com/pub/totalcontrol-hubs/tcmwin/

At 07:07 PM 1/30/98 -0400, James Wilson wrote: >Anyone know where I can get the latest release. The website is not ftp'ing >correctly and I need this STAT. > >James >james@cruxnet.com > I do but if I told you, I'd have to kill you.. :) Call tech support. They might tell you.. Or at least email it to you. TIP: Use "Teleport" or other webcrawler to download the entire site when they get it working again. Then burn a CD so you can browse the site offline.. Teleport lets you update files so you can keep up to date once a month or so.. It's the only way to go.. It's fun to click on a download.. They go at harddisk (or CD) speeds.. 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

At 03:19 PM 1/30/98 -0800, MegaZone wrote: >I'm not going to get into a feature-by-feature comparision, especially >on an unreleased product, but what in particular is unforgivably missing >tin your mind? From what I've seen, and as you told me that I have not see it all, I would want the interface to be as user friendly as TCMs. To be quite honest, it is hard for me to compare since I am so used to TCM. But from my perspective, Amber looks really unprofessional compared to it. It looks like what it is, a freshly made product that still needs a lot of work on it and improvement. > >>I understand you are big lover of Portmasters(after all you work for >>Livingston). However, I, and I am sure a lot of other folks here, do not >>appreciated the fact that you always stick your 5c into every conversation >>where you find you can promote livingston products here. THIS IS NOT A > >Frankly I don't care. And, as you know, I have 2 weeks left here before >I start at GTE/BBN. And I'll be on from my personal domain after that to >be sure - and I'll be less restrained than ever. You don't like it? Filter >me out - filter, procmail, or whatever you prefer. The technology exists, >so use it instead of complaining. This is not for my personal satisfaction. I'll block your email if I would so desire. I am just saying that at times you are making a lot of people feel really bad when they are stating problems they are having on this list, and you stick your nose into it and say: "Well you know what, Livingston has it already." If USR/3com was in the same business, then every time anyone on portmaster-users mentioned MLPPP, they'd respond: "Well you know what, 3com had this since heck knows when." Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

At 02:52 PM 1/30/98 -0800, Josh Richards wrote: >Garry, aren't you on the portmaster-users list? > Of course I am. I was also the one begging for a 3.8 with Amber so I can have better control of my PM3 here. :)) Garry Shtern shterng@akula.com Chief Network Administrator http://www.akula.com Akula Communications Corp. tel. (212) 292-8892

Anyone know where I can get the latest release. The website is not ftp'ing correctly and I need this STAT. James james@cruxnet.com

Is the telco sending you timing on that circuit? -CH On Fri, 30 Jan 1998, James Wilson wrote: > It seems that after about 15-30 minutes of being up, our pri get out of sync > with the hub. All the calls either get busies or nothing picks up. A > hardware reset of the card restores the pri. Any thoughts on the solution? > > 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. >

At 11:15 AM 1/29/98 -0700, Pete Ashdown wrote: >Allen Marsalis said once upon a time: > >>PROBLEM: >>Some users can't connect.. They receive the following error.. >>This happens on both arc equipped and netserver equipped chassis.. > >I would go through *ALL* the settings on your HDM's and verify they are set >to the defaults, except for the ISDN settings I posted to the list the >other day. > >I had some HDMs shipped to me configured to connect at only 300 BPS!! The >problem drove me crazy until I went through all the settings and verified >they were correct. > >The HDM template makes this fairly easy. > Well it looks like re-flashing the hdms and hiperarc last night did the trick! Thanks krish, pete, and whoever the person was who said this happened to them and fixed it by re-flashing. This is still a mystery in that I reflashed the hdms with the same code version they reported having to start with.. We also changed the "selective reject" parameter on one hdm so maybe that had something to do with it. But now that things are normal, I can kick back and play some lag-free Quake! 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

> From: rpanula@dacmail.net (Russ Panula) > >>Anyone know where I can get the latest release. The website is not ftp'ing >>correctly and I need this STAT. >> >>James >>james@cruxnet.com > > >ftp://username:password@totalservice.usr.com/pub/totalcontrol-hubs/tcmwin/ This is actualy a sim link back to /pub/.files/<your file here>. And those files are returning "file not found" even though they are listed in the directory. That is why I was asking if anyone had the code. But I talked to a tech and after some prodding, he is emailing me the code. James

Well just when all was going well. One hdm is empty while we are having busys.. OK so I do a "display atstat" and the Modem Connect column shows N/A for all channels. Doing a reboot or a sinserv didn't help. I don't want to reboot the hiperarc if I don't have to. Anyone got a quick answer on a Friday night??.. allen

At 11:14 AM 1/30/98 -0600, Tatai SV Krishnan wrote: >On Fri, 30 Jan 1998, Allen Marsalis wrote: > >> Well just when all was going well. One hdm is empty while we are >> having busys.. OK so I do a "display atstat" and the Modem Connect >> column shows N/A for all channels. Doing a reboot or a sinserv >> didn't help. I don't want to reboot the hiperarc if I don't have >> to. Anyone got a quick answer on a Friday night??.. >> >> allen > >On the Hiper arc do a list interface, it will show if the modems are >owned by the Hiper arc or not. If its not owned by hiper then you have >to change it by using the command set cha sl <slot#> ow y > >krish > As it turns out, the hdm was set for 5ESS and not DMS100.. I thought I checked all the hdms after flashing them but I guess I missed it. Dunno why the configuration would change all by itself. Previously it was configured and working for days with the hiperarc.. But I've heard how sometimes re-flashing does strange things to the config.. One positive note. I can't believe how easy it is to configure a hdm. And the doc was so clear and simple.. I am now very glad that I chose to go Hiper this soon. And quake is *alot* better too. Now I just hope things are stable and I will once again be a very satisfied usr customer.. 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

Was having problems with ISDN Dial up. Thanks to all who replied to me, but nothing worked... Bell Atlantic could not find a fault. Out of despiration, I powered off the chassis, and when it came back up it worked... Hated to do that, but sometimes we have no choice... 73's John Campbell - KC4LWI Owner - Roanoke Virginia Net http://www.roava.net mailto:sparky@roava.net

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

Is it possible to get a list of users that are login via SNMP? I didn't find any obvious answers glancing through the SNMP files. Should this be obtained through NMC or the Netserver? Thanks, Tim

> From: Charles Hill <chill@ionet.net> > >Is the telco sending you timing on that circuit? -CH I assumed so. But after much hair pulling I tried setting the internal timing as high and the span as medium. This has worked so far. Can anyone see a problem with doing this? After all, it's working. >On Fri, 30 Jan 1998, James Wilson wrote: > >> It seems that after about 15-30 minutes of being up, our pri get out of sync >> with the hub. All the calls either get busies or nothing picks up. A >> hardware reset of the card restores the pri. Any thoughts on the solution? >> >> James >> James

Didn't know whether to send this to the list as a whole or to MZ directly....guess I'll default to the list. :/ Some personal comments, so skip if you feel like it.  Thus spake MegaZone >>I understand you are big lover of Portmasters(after all you work for >>Livingston). However, I, and I am sure a lot of other folks here, do not >>appreciated the fact that you always stick your 5c into every conversation >>where you find you can promote livingston products here. THIS IS NOT A Personally, I *like* having reps from other companies here. Let's me keep up with other equipment without having to keep up with thier lists as much. Frankly, I don't have time to be on portmaster-users to support the 3 PM-25's that we have (and basically don't ever touch), but it is nice to hear what Livingston^WLucent is doing with the PM stuff some. >Frankly I don't care. And, as you know, I have 2 weeks left here before >I start at GTE/BBN. Wha? Did I miss this, or was it posted to another list? In any case, good luck at GTE/BBN. >And I'll be on from my personal domain after that to >be sure - and I'll be less restrained than ever. You mean you were being restrained? ;) Just kidding, I've been quite surprised at how quiet you've been around these parts. :) >You don't like it? Filter >me out - filter, procmail, or whatever you prefer. The technology exists, >so use it instead of complaining. Possible buffer overflows in filter...better use procmail. :) Though I haven't made the switch yet...just too busy. >Again - if you don't like it, use the technology and just filter my email >out. Simple. I've always laughed at anyone who whines about my posting, >because it is their own damn fault that they read it in the first place. I've always wondered about this as well. People complaining about certain posts and messages in varius fora when the technology to send those to the ol' bit bucket in the sky is extremely simple and available. -- Jeff McAdams Email: jeffm@iglou.com Chief Network Administrator Voice: (502) 966-3848 IgLou Internet Services (800) 436-4456