Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket. There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?) HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho). At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
--- "Don't go with a spineless ISP; we have more backbone." Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc... blech. I would not recommend them to my worst enemy. Charles ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc...
blech. I would not recommend them to my worst enemy.
Charles
~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com
On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
Why wouldn't upgrading the router's CPU or splitting the load between two separate routers solve the problem? Maybe that's not an overnight solution [unless you have spare RSPs in the POP/colo point] but it seems pretty straight forward to me. -Deepak. On Mon, 30 Jun 1997, Wayne Bouchard wrote:
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc...
blech. I would not recommend them to my worst enemy.
Charles
~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com
On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
Thats my point. It often ISN'T easy. For example, say you've got a 7505 in one spot and find its getting killed. You can't put an rsp4 in a 7505. You've got to replace the router. Anyone got any spare 7507s or 7513s lying around? Chances are that if they do, its 2 or 3. Its actually going to be a bit more problematic to split the load between two routers since that can mean notable changes in the internal routing or the BGP tables. Not to mention calculating what kind of traffic is going to pass BETWEEN the two routers. Straightforward, yes. Easy, no.
Why wouldn't upgrading the router's CPU or splitting the load between two separate routers solve the problem? Maybe that's not an overnight solution [unless you have spare RSPs in the POP/colo point] but it seems pretty straight forward to me.
-Deepak.
On Mon, 30 Jun 1997, Wayne Bouchard wrote:
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc...
blech. I would not recommend them to my worst enemy.
Charles
~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com
On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
Umm.. 7505, 7507 and 7513s all use the same processors.. RSP4s do work in 7505s. rob
Thats my point. It often ISN'T easy. For example, say you've got a 7505 in one spot and find its getting killed. You can't put an rsp4 in a 7505. You've got to replace the router. Anyone got any spare 7507s or 7513s lying around? Chances are that if they do, its 2 or 3.
Its actually going to be a bit more problematic to split the load between two routers since that can mean notable changes in the internal routing or the BGP tables. Not to mention calculating what kind of traffic is going to pass BETWEEN the two routers.
Straightforward, yes. Easy, no.
Why wouldn't upgrading the router's CPU or splitting the load between two separate routers solve the problem? Maybe that's not an overnight solution [unless you have spare RSPs in the POP/colo point] but it seems pretty straight forward to me.
-Deepak.
On Mon, 30 Jun 1997, Wayne Bouchard wrote:
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc...
blech. I would not recommend them to my worst enemy.
Charles
~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com
On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
On Mon, 30 Jun 1997 18:17:48 -3100 (MST), web@typo.org writes:
Thats my point. It often ISN'T easy. For example, say you've got a 7505 in one spot and find its getting killed. You can't put an rsp4 in a 7505. You've got to replace the router. Anyone got any spare 7507s or 7513s lying around? Chances are that if they do, its 2 or 3.
And yet.. you folks keep on using Cisco.. <duck> -Jon ----------------------------------------------------------------- * Jon Green * "Life's a dance * * jcgreen@netINS.net * you learn as you go" * * Finger for Geek Code/PGP * * * #include "std_disclaimer.h" * http://www.netins.net/showcase/jcgreen * -------------------------------------------------------------------------
On Mon, 30 Jun 1997 21:26:18 -0500 Jon Green <jcgreen@netins.net> wrote:
And yet.. you folks keep on using Cisco..
I agree Jon its insane, I chuckle when I see all these networks melting because the insist on using Cisco routers. Vote with your money and buy something that works. Cheers, Neil. -- Neil J. McRae. Alive and Kicking. Domino: In the glow of the night. neil@DOMINO.ORG NetBSD/sparc: 100% SpF (Solaris protection Factor) Free the daemon in your <A HREF="http://www.NetBSD.ORG/">computer!</A>
What would you suggest? I've looked into the Ascend GRF, and if it performs as much as they hype it (never trust a salesman) then it must be one kickbutt little router. My only question is who has some in a real world setup, doing BGP/OSPF/RIP? I tried talking to psi.net about their GRF switchover, and ended up with someone trying to sell me bandwidth. Joe Shaw - jshaw@insync.net NetAdmin - Insync Internet Services "Learn more, and you will never starve." - Paraphrase of Lee On Tue, 1 Jul 1997, Neil J. McRae wrote:
On Mon, 30 Jun 1997 21:26:18 -0500 Jon Green <jcgreen@netins.net> wrote:
And yet.. you folks keep on using Cisco..
I agree Jon its insane, I chuckle when I see all these networks melting because the insist on using Cisco routers. Vote with your money and buy something that works.
Cheers, Neil. --
On Tue, 1 Jul 1997 10:28:14 -0500 (CDT) Joe Shaw <jshaw@insync.net> wrote:
What would you suggest? I've looked into the Ascend GRF, and if it performs as much as they hype it (never trust a salesman) then it must be one kickbutt little router. My only question is who has some in a real world setup, doing BGP/OSPF/RIP? I tried talking to psi.net about their GRF switchover, and ended up with someone trying to sell me bandwidth.
I'm using the GRF-400. It has multiple 155M SDH ATM links in it, works like a dream... I'm awaiting an FDDI link in to the LINX and then it will be peering with around 25 other LINX members, so far in testing its done this with its eyes closed... # uname -a Ascend Embedded/OS the-count.router.COLT.NET 2.1 Ascend Embedded/OS GR TA1.3.5 Kernel #1 (nit): Wed Apr 23 07:59:43 CDT 1997 nit@bsd2.netstar.com:/nit/A1_3_5/BSDI/sys/compile/RMS i386 GateD-the-count.router.COLT.NET> show bgp peeras 6453 1GRF group type External AS 6453 local 8220 flags <> 1GRF Peer: 207.45.206.245 ID: 207.45.206.253 Version: 4 Gateway: (null) 1GRF Local ID: 195.110.64.2 1GRF Local Addr: 1GRF 207.45.206.246+1540 1GRF flags 0x20 1GRF state 0x6 <Established> 1GRF options 0x400001 <MetricOut MED> 1GRF group bit: 0 1GRF MED exported 1 1GRF proto-precedence/BGP preference 170/- 1GRF messages in 585246 (updates 573347, not updates 11899) 36927948 octets 1GRF messages out 11908 (updates 2, not updates 11906) 226348 octets 1GRF Last traffic (seconds): Received 17, Sent 44, Checked 44 1GRF Inbound Timer: 0 1GRF Outbound Timer: 0 1GRF Received and buffered Octets: 0 1GRF Active Holdtime: 180 1GRF Route Queue Timer: 1GRF unset 1GRF Route Queue: empty 1GRF Peer: 207.45.199.245 ID: 207.45.199.253 Version: 4 Gateway: (null) 1GRF Local ID: 195.110.64.2 1GRF Local Addr: 1GRF 207.45.199.246+179 1GRF flags 0x20 1GRF state 0x6 <Established> 1GRF options 0x400011 <MetricOut Preference MED> 1GRF group bit: 0 1GRF MED exported 2 1GRF proto-precedence/BGP preference 180/- 1GRF messages in 138687 (updates 136013, not updates 2674) 8658107 octets 1GRF messages out 2676 (updates 1, not updates 2675) 50891 octets 1GRF Last traffic (seconds): Received 16, Sent 18, Checked 18 1GRF Inbound Timer: 0 1GRF Outbound Timer: 0 1GRF Received and buffered Octets: 0 1GRF Active Holdtime: 180 1GRF Route Queue Timer: 1GRF unset 1GRF Route Queue: empty -- Neil J. McRae. Alive and Kicking. Domino: In the glow of the night. neil@DOMINO.ORG NetBSD/sparc: 100% SpF (Solaris protection Factor) Free the daemon in your <A HREF="http://www.NetBSD.ORG/">computer!</A>
On Tue, 1 Jul 1997, Neil J. McRae wrote:
On Tue, 1 Jul 1997 10:28:14 -0500 (CDT) Joe Shaw <jshaw@insync.net> wrote:
What would you suggest? I've looked into the Ascend GRF, and if it performs as much as they hype it (never trust a salesman) then it must be one kickbutt little router. My only question is who has some in a real world setup, doing BGP/OSPF/RIP? I tried talking to psi.net about their GRF switchover, and ended up with someone trying to sell me bandwidth.
I'm using the GRF-400. It has multiple 155M SDH ATM links in it, works like a dream...
Yes the ATM card is nice, so far we have over 138 external peering session without problem. Nathan Stratton President, NetRail,Inc. ------------------------------------------------------------------------ Phone (888)NetRail NetRail, Inc. Fax (404)522-1939 230 Peachtree Suite 500 WWW http://www.netrail.net/ Atlanta, GA 30303 ------------------------------------------------------------------------ "No king is saved by the size of his army; no warrior escapes by his great strength. - Psalm 33:16
The ascend GRF is one kickbutt little router. We are only using it for static routing right now, because thats all that we need it for, but it does have the capabilities to do ospf, bgp and rip. We switched from a cisco 7500 because of all of the problems that we were having with packet loss and speed issues and haven't had a problem with it in the last 2 1/2 months that we've been using it. Stephen Dolloff (System Administrator - McHenryCom) Joe Shaw wrote:
What would you suggest? I've looked into the Ascend GRF, and if it performs as much as they hype it (never trust a salesman) then it must be one kickbutt little router. My only question is who has some in a real world setup, doing BGP/OSPF/RIP? I tried talking to psi.net about their GRF switchover, and ended up with someone trying to sell me bandwidth.
Joe Shaw - jshaw@insync.net NetAdmin - Insync Internet Services "Learn more, and you will never starve." - Paraphrase of Lee
On Tue, 1 Jul 1997, Neil J. McRae wrote:
On Mon, 30 Jun 1997 21:26:18 -0500 Jon Green <jcgreen@netins.net> wrote:
And yet.. you folks keep on using Cisco..
I agree Jon its insane, I chuckle when I see all these networks melting because the insist on using Cisco routers. Vote with your money and buy something that works.
Cheers, Neil. --
On Tue, 1 Jul 1997, Joe Shaw wrote:
What would you suggest? I've looked into the Ascend GRF, and if it performs as much as they hype it (never trust a salesman) then it must be one kickbutt little router. My only question is who has some in a real world setup, doing BGP/OSPF/RIP? I tried talking to psi.net about their GRF switchover, and ended up with someone trying to sell me bandwidth.
I am happy with them we have 14 in our network right now. Yes they have some small problems, but Ascend GRF tech support is the best I have seen. We have some new Cisco stuff on the way, and in a few weeks will know if we are going to ditch GRF and with with Cisco. So far we are still 100% cisco free, but that may change in a few weeks. Try the GRF, it is a nice box. Nathan Stratton President, NetRail,Inc. ------------------------------------------------------------------------ Phone (888)NetRail NetRail, Inc. Fax (404)522-1939 230 Peachtree Suite 500 WWW http://www.netrail.net/ Atlanta, GA 30303 ------------------------------------------------------------------------ "No king is saved by the size of his army; no warrior escapes by his great strength. - Psalm 33:16
On Tue, 1 Jul 1997, Nathan Stratton wrote:
I am happy with [the GRF] we have 14 in our network right now. Yes they have some small problems, but Ascend GRF tech support is the best I have seen. We have some new Cisco stuff on the way, and in a few weeks will know if we are going to ditch GRF and with with Cisco. So far we are still 100% cisco free, but that may change in a few weeks.
Maybe they should send some of those tech support people over to work on the Ascend Max code. 9 weeks now of having a totally reproducable routing loss bug into them, and they won't even bother returning email or call-backs to me anymore. The previous code revisions made the boxes reboot at random times. We picked the lesser of two evils..
At 02:41 PM 7/1/97 -0500, Doug McIntyre wrote:
On Tue, 1 Jul 1997, Nathan Stratton wrote:
I am happy with [the GRF] we have 14 in our network right now. Yes they have some small problems, but Ascend GRF tech support is the best I have seen. We have some new Cisco stuff on the way, and in a few weeks will know if we are going to ditch GRF and with with Cisco. So far we are still 100% cisco free, but that may change in a few weeks.
Maybe they should send some of those tech support people over to work on the Ascend Max code. 9 weeks now of having a totally reproducable routing loss bug into them, and they won't even bother returning email or call-backs to me anymore. The previous code revisions made the boxes reboot at random times. We picked the lesser of two evils..
GRF tech support is the *same* group of people as those you refer to here. ....I've asked someone to look into your reported troubles and get back to you/me on why there has been no further contact. Kevin
On Mon, 30 Jun 1997, Wayne Bouchard wrote: ==>Thats my point. It often ISN'T easy. For example, say you've got a ==>7505 in one spot and find its getting killed. You can't put an rsp4 in ==>a 7505. You've got to replace the router. Anyone got any spare 7507s ==>or 7513s lying around? Chances are that if they do, its 2 or 3. You *can* put an RSP4 in a 7505. Only 1 CyBus is used on the card, but it works just fine. /cah
On Mon, 30 Jun 1997 18:17:48 -3100 (MST) Wayne Bouchard <web@typo.org> wrote:
Thats my point. It often ISN'T easy. For example, say you've got a 7505 in one spot and find its getting killed. You can't put an rsp4 in a 7505. You've got to replace the router. Anyone got any spare 7507s or 7513s lying around? Chances are that if they do, its 2 or 3.
Why _must_ one use a Cisco for this? -- Neil J. McRae. Alive and Kicking. Domino: In the glow of the night. neil@DOMINO.ORG NetBSD/sparc: 100% SpF (Solaris protection Factor) Free the daemon in your <A HREF="http://www.NetBSD.ORG/">computer!</A>
On Mon, 30 Jun 1997, Wayne Bouchard wrote:
Thats my point. It often ISN'T easy. For example, say you've got a 7505 in one spot and find its getting killed. You can't put an rsp4 in a 7505. You've got to replace the router. Anyone got any spare 7507s or 7513s lying around? Chances are that if they do, its 2 or 3. It's often not easy to FIX. But it's not hard to _notify the customers_. And to work around by some way (temporary).
My first question is: Have you ever tried putting an RSP2 in a 7505? It ships with an RSP, second, have you tried putting an RSP4 in a 7505? It works too. Cisco isn't real happy about it, and of course I wouldn't recommend it, but it works. Next, if your network is that busy, you honestly don't keep that kind of spare/upgrade equipment around? -Deepak. On Mon, 30 Jun 1997, Wayne Bouchard wrote:
Thats my point. It often ISN'T easy. For example, say you've got a 7505 in one spot and find its getting killed. You can't put an rsp4 in a 7505. You've got to replace the router. Anyone got any spare 7507s or 7513s lying around? Chances are that if they do, its 2 or 3.
Its actually going to be a bit more problematic to split the load between two routers since that can mean notable changes in the internal routing or the BGP tables. Not to mention calculating what kind of traffic is going to pass BETWEEN the two routers.
Straightforward, yes. Easy, no.
Why wouldn't upgrading the router's CPU or splitting the load between two separate routers solve the problem? Maybe that's not an overnight solution [unless you have spare RSPs in the POP/colo point] but it seems pretty straight forward to me.
-Deepak.
On Mon, 30 Jun 1997, Wayne Bouchard wrote:
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc...
blech. I would not recommend them to my worst enemy.
Charles
~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com
On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
Why wouldn't upgrading the router's CPU or splitting the load between two separate routers solve the problem? Maybe that's not an overnight solution [unless you have spare RSPs in the POP/colo point] but it seems pretty straight forward to me.
-Deepak.
Probably because the CPU that's overloaded is not the primary Cisco processor, but an RSP board or somesuch. When you have customers coming into Frame Relay switches, it's not so easy to just move customers between HSSI's without just overloading another router. Steve Mansfield steve@nwnet.net NorthWestNet Network Engineer 425-649-7467
Cheap Cheap Cheap Cheap Cheap, that's all I can figure. No budget? As a customer, I feel really shafted considering the rates that UUNet charges. As for the time element, anyone off of this router could have seen this coming, oh, about 8 months ago... In the future, I'd like to hand my (well, not MY, the owner's) thousands to someone that understands the word proactive. Not being big enough to get a fully-routable CIDR block kind of locks one in from making snap decisions on bandwidth. My UUNet complaints file contains at least 100 messages with examples of the poor performance of that router. I get form letters in return... "Look, I'm pinging you at 3am, and look how fast it goes!". blech blech blech. Don't use UUNet if you are not a multihomed, CIDR-ized playah. C ~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com On Mon, 30 Jun 1997, Deepak Jain wrote:
Date: Mon, 30 Jun 1997 21:15:07 -0400 (EDT) From: Deepak Jain <deepak@jain.com> To: Wayne Bouchard <web@typo.org> Cc: Charles Sprickman <spork@inch.com>, alex@nac.net, jbeck@connectnet.com, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
Why wouldn't upgrading the router's CPU or splitting the load between two separate routers solve the problem? Maybe that's not an overnight solution [unless you have spare RSPs in the POP/colo point] but it seems pretty straight forward to me.
-Deepak.
On Mon, 30 Jun 1997, Wayne Bouchard wrote:
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
Oh God, NYC1. Look out for that ugly beast... We are unfortunate enough to have a connection off of that thing. The NOC is good, if you're up for entertainment. Let me count the times I've heard "what's traceroute" on the phone with them. Let me tell you about outages where they "forgot to take the loopback off the T" for like 6 hours. Or my favorite (once you hit a high-level brainwashed engineer) the good old "I don't see *any* packet loss, so everything is OK". "Yes, but I'm seeing ping times of 500ms to your router, and our line isn't near saturation by a long shot"... "I don't see any packet loss", etc...
blech. I would not recommend them to my worst enemy.
Charles
~~~~~~~~~ ~~~~~~~~~~~ Charles Sprickman Internet Channel INCH System Administration Team (212)243-5200 spork@inch.com access@inch.com
On Mon, 30 Jun 1997, Alex Rubenstein wrote:
Date: Mon, 30 Jun 1997 18:18:06 -0400 From: Alex Rubenstein <alex@nac.net> To: Josh Beck <jbeck@connectnet.com>, nanog@merit.edu Subject: Re: Plethora of UUnet outages and instabilities
All I know is they suck moose nipples, and I disconnected from them last week. They have been having problems in/around cr1.nyc1 for MONTHS now, and they are doing nothing (that I can see) to fix it. Thier first-level support is horrendous. One guy once said to me, "Routers have processors?!". Another hung up when I told him to put "CUSTOMER IS IRATE" on the trouble ticket.
There are some GOOD guys over there, but them seem to be disappearing? (are you there, Mr. Hannan?)
HELLO, GENUITY (anyone who is looking for a new provider, Genuity is DAMN good. and a DAMN good NOC. Sales guys are a little whacky, tho).
At 02:12 PM 6/30/97 -0700, Josh Beck wrote:
Has everyone else been seeing the almost daily UUnet outages recently, does anyone know what the true causes of these have been?
---
"Don't go with a spineless ISP; we have more backbone."
Alex Rubenstein -- alex@nac.net -- KC2BUO -- www.nac.net net @ccess corporation, 201-983-0725 -- 201-983-0725
---------------------------------------------------------------------- Wayne Bouchard GlobalCenter web@primenet.com Primenet Network Engineering Internet Solutions for (602) 416-6422 800-373-2499 x6422 Growing Businesses FAX: (602) 416-9422 http://www.primenet.com http://www.globalcenter.net ----------------------------------------------------------------------
On Mon, 30 Jun 1997, Wayne Bouchard wrote:
I know one of the problems that UUNET has been having (and I believe several other folks as well) is high CPU usage on many of their routers. (When the ping times to the router itself get high its either a congested link or, usually, a router thats spinning hard..) However, thats not a problem you can fix easily or fix overnight.
No, as a matter of fact, it's not, as it took UUNet months to upgrade a Net Edge box that they were originaly going to replace. It took them months to eventually get around to upgrading, and then it took another upgrade later in the week to fix the previous upgrade. UUNet has far too much backing and resources to be having these kinds of problems. UUNet's NOC is always fun to deal with. I spent the better part of two days trying to find out why the bgp tables we were getting from them were so screwed after they rebooted routers in 3 different states. The rebooted them again 2 days later, and the problems disappeared. The whole time, I've got engineers telling me that the setup that's worked flawlessly since I've been here has got to be what's wrong. Then, the next day, I read on here that everyone is having UUNet problems. Confront the guy at UUNet again, and he admits they are having BGP problems in several states. An honest answer would be nice. Also, they don't make it part of their known outages that their news system is currently overloaded and won't allow customers to feed articles to them. I made it a point to check, and the current status of the mail and news was posted as normal. I get on the phone with the person in charge of news, and she says it's a capacity issue that they've been dealing with for the last week or so, and they're about to have a meeting on it right away. Once again, I got an honest answer, but only after being lied to. I was once a strong UUNet supporter, but their performance and quality of service needs to improve drasticly. I hope they do... Joe Shaw - jshaw@insync.net NetAdmin - Insync Internet Services Learn more, and you will never starve.
participants (15)
-
Alex P. Rudnev -
Alex Rubenstein -
Charles Sprickman -
Craig A. Huegen -
Deepak Jain -
Doug McIntyre -
Joe Shaw -
Jon Green -
Kevin Smith -
Nathan Stratton -
Neil J. McRae -
Robert Bowman -
Stephen Dolloff -
Steve Mansfield -
Wayne Bouchard