- NANOG - lists.nanog.org

clinton creates internet study group (and police force?)
by Ehud Gavron 09 Aug '99

09 Aug '99

text of the EO: http://www.usnewswire.com/topnews/Current_Releases/0807-107.htm One analysis: http://www.worldnetdaily.com/bluesky_btl/19990809_xcbtl_clinton_co.shtml Ehud

1 0

Re: What frame relay switch is causing MCI/Worldcom such grief?
by Vadim Antonov 09 Aug '99

09 Aug '99

> So I guess that means you know more than all those smart people (on > both the service provider side and equipment side) who are helping > incorporate lessons from Frame Relay and ATM circuit routing into > MPLS? Yep. I can put together two and two and get four. I'm not going to read yet another lecture on computational complexity of adaptive circuit routing but so far nobody figured out a way to make it work in a network comparable to the current Internet in size. That's why nobody is even attempting to do MPLS or whatever across IXPs and only do in in the interior (actually this is also misguided, but for different reasons, having more to do with implementation complexity). BTW, a good many smart people who actually built the biggest chunks of the Internet backbone and the software which keeps it running share my opinion. In any case, the argument that million lemmings cant be all wrong is particularly hilarious. The issue of virtual circuit vs per-hop routing was discussed extensively in this forum years ago (in ATM vs IP context) and the consensus was pretty much against ATM. I would advise you to review the archives and see for yourself. Since there were no dramatic advances in either understanding of the problem, circuit-routing algorithms, or basic computing technology (well, the quantum computing can concievably solve the computablibity problem for circuit routing, but not any time soon :) i see little reason to start the arguments all over again. Of course, you may have something really new to say on the topic, but i would check that that wasn't already beaten to death. Regards, --vadim

3 2

Re: What frame relay switch is causing MCI/Worldcom such grief?
by smd＠clock.org 09 Aug '99

09 Aug '99

| So I guess that means you know more than all those smart people (on | both the service provider side and equipment side) who are helping | incorporate lessons from Frame Relay and ATM circuit routing into | MPLS? Yes. Sean.

2 1

Re: What frame relay switch is causing MCI/Worldcom such grief?
by Vadim Antonov 09 Aug '99

09 Aug '99

Prabhu -- a particluar implementation does not matter when the underlying idea is bad. Any circuit routing for data is broken-as-designed. (BTW, the irrelevance of the particulars is why i omitted attribution in the first place. This was a comment on the general state of affairs, not on the situation. Thanks for providing details, though). --vadim >From: Prabhu Kavi <prabhu_kavi(a)tenornetworks.com> >Vadim, >Your response is unrelated to the message I replied to, which >asked about the link state algorithm that Cascade uses. >As far as I know, nobody has announced which set of MCI Wordlcom's >switches were responsible for their outage. Since MCI Worldcom >uses multiple types in their network, statements like "goes as >predicted - complexity bites" is irresponsible at best, even if it >is later determined the Cascade switches were actually the cause >of the outage. >Prabhu

2 1

OUTAGE: MCI/Worldcom frame-relay network
by Sean Donelan 09 Aug '99

09 Aug '99

MCI/Worldcom's frame-relay network has been experiencing serious problems for the last six hours or so. Originally thought to be an upgrade gone bad, now the cause is unknown. MCI/Worldcom is in the process of shutting down and rebuilding different switches one-by-one. It is unknown when the network will stabilize. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

2 1

RE: What frame relay switch is causing MCI/Worldcom such grief?
by Baird, Alan 09 Aug '99

09 Aug '99

I believe WorldCom is using the BSTDX-9000/CBX-500 switches on their network. The problems that they are having sound suspiciously like what would happen to a network when it exceeded the 200 node limit on some of Ascend's older code. Perhaps they haven't got around to going to the new code release? Alan Baird Williams Network Architecture Lab -----Original Message----- On Mon, 9 Aug 1999, Alan Hannan wrote: What frame relay switch is causing MCI/Worldcom such grief?

1 0

Re: What frame relay switch is causing MCI/Worldcom such grief?
by Christopher Gibiault 09 Aug '99

09 Aug '99

current resume attached -----Original Message----- From: Prabhu Kavi <prabhu_kavi(a)tenornetworks.com> To: Vadim Antonov <avg(a)kotovnik.com> Cc: nanog(a)merit.edu <nanog(a)merit.edu> Date: Monday, August 09, 1999 2:52 PM Subject: Re: What frame relay switch is causing MCI/Worldcom such grief? > >Vadim, > >Your response is unrelated to the message I replied to, which >asked about the link state algorithm that Cascade uses. > >As far as I know, nobody has announced which set of MCI Wordlcom's >switches were responsible for their outage. Since MCI Worldcom >uses multiple types in their network, statements like "goes as >predicted - complexity bites" is irresponsible at best, even if it >is later determined the Cascade switches were actually the cause >of the outage. > >Prabhu > >Vadim Antonov wrote: >> >> This pretty much goes as predicted - complexity bites. Watch for more. >> >> What happened to the KISS principle? >> >> --vadim >> >> >On top of OSPF, there is a protocol which is responsible for actually >> >setting up the connections according to specific constraints (e.g. >> >bandwidth, QoS level). Again, this is similar to the work >> >now being done in working groups responsible for CR-LDP and RSVP. >> >Cascade released its version in 1993, and has been refining it >> >ever since. > >---------------------------------------------------------------------- >Prabhu Kavi Phone: 978-264-4900 x125 >Director, Prod. Mgmt. FAX: 978-264-0671 >Tenor Networks Email: prabhu_kavi(a)tenornetworks.com >50 Nagog Park WWW: www.tenornetworks.com >Acton, MA 01720 >---------------------------------------------------------------------- > >

1 0

Re: What frame relay switch is causing MCI/Worldcom such grief?
by Vadim Antonov 09 Aug '99

09 Aug '99

This pretty much goes as predicted - complexity bites. Watch for more. What happened to the KISS principle? --vadim >On top of OSPF, there is a protocol which is responsible for actually >setting up the connections according to specific constraints (e.g. >bandwidth, QoS level). Again, this is similar to the work >now being done in working groups responsible for CR-LDP and RSVP. >Cascade released its version in 1993, and has been refining it >ever since.

2 1

CORRECTION/UPDATE: Re: OUTAGE: MCI/Worldcom frame-relay network
by Sean Donelan 09 Aug '99

09 Aug '99

UUNET has restored the frame-relay outage paragraph to their web page. It was gone at 12:00EDT (Noon), it was back at 12:29EDT. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Re: OUTAGE: MCI/Worldcom frame-relay network
by Sean Donelan 09 Aug '99

09 Aug '99

alan(a)globalcenter.NET (Alan Hannan) writes: > PS - sean's messages were not 'highly accurate' as they maintained > that WCOM was not reporting on the problems, while they were. To be more precise Alan, Sean said MCI/Worldcom's web site continued not to have any information about the outage. That is still true. I did miss the reference on the UUNET site, but now even UUNET has removed reference to the frame-relay outage from their web site. MCI/Worldcom's web site is reporting on their new 5cent a minute long distance plan this morning, so I suppose that means MCI/Worldcom's PR people have returned to work. But nothing about their frame-relay network on the website. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0