- Test - lists.nanog.org

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

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

09 Aug '99

> The information displayed here is not always accurate, and > often subjective. The information displayed in many forums is often subjective and not always accurate. Obviously the MCI/Worldcom spokepeople were overly optimistic in the CBS Marketwatch article, and also the Reuters article, before they left on Friday. As I mentioned the longer the outage has continued the fewer people seem to work at MCI/Worldcom. > Usually I'd argue that any information is better than no > information, but when the quality is of suspect value, I'm > not sure the viewpoint stands. Sorry Alan, I didn't realize that. I won't post any more about this outage. Call 800-644-NEWS if you want to hear what's happening with MCI/Worldcom. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

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

09 Aug '99

SEAN(a)dra.COM (Sean Donelan) writes: >MCI/Worldcom's web site continues its long silence about this outage. A person pointed out to me that UUNET (part of MCI/Worldcom) has a note about the outage on their web site. Obviously not the first place I would have looked, but nevertheless a statement on a MCI/Worldcom web site. -- 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 08 Aug '99

08 Aug '99

SEAN(a)dra.COM (Sean Donelan) writes: >But no rest for the exhausted. There is a software upgrade scheduled later >tonight. I'm trying to post no more than once in 12 hours. The attempt to load software last night did not go well. The network collapsed again. If it didn't work the first three times, try, try, try again. MCI/Worldcom is resetting cards and switches one more time. In an interesting reverse escalation, the longer the outage continues; the fewer people there seem to be at worldcom. The VPs and managers seem to have vanished, leaving behind call takers and an occasional engineer who accidently picks up a phone. -- 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 08 Aug '99

08 Aug '99

SEAN(a)dra.COM (Sean Donelan) writes: >It is unknown when the network will stabilize. MCI/Worldcom now claims the network is 'up.' However it is suffering heavy congestion, in some parts dropping 90%-100% of the frames. We're debating with MCI/Worldcom what a 'working' PVC really means. But no rest for the exhausted. There is a software upgrade scheduled later tonight. I thought everyone in this business knew who Bernie was, I was wrong. Bernard J. Ebbers is the CEO of MCI/Worldcom. No I don't expect Bernie to know what a frame-relay switch looks like. But when 10% of the Chicago Board of Trade goes down the tubes, you expect the CEO of the network provider to be nominally in charge. Ebay's Meg Whitman showed up on CNBC and had public statements, AT&T's Michael Armstrong kept customers updated during their various outages which were much shorter than MCI/Worldcom's. Look on MCI/Worldcom's web site and the most recent statement you'll find is a press release about second quarter earnings. Hint: never sign a five-year contract with a provider. You never know just how bad they will get. -- 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 07 Aug '99

07 Aug '99

SEAN(a)dra.COM (Sean Donelan) writes: >It is unknown when the network will stabilize. Well, we still don't have all our PVC's back. We were back to the point of having only one PVC down, when the entire MCI/Worldcom frame-relay network collapsed again last night. Supposedly its been escalated to Bernie now. MCI/Worldcom's frame-relay network continues to 'toggle' switches throughout the network. There was an OC12 outage in Riverdale causing congestion. Several cards no longer worked at all. MCI/Worldcom is working on reloading software on 54 switches. Still no ETR. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0