Sorry for this late followup. This thread was just brought to my attention.
Sounds like the cisco bug which has affected several different providers over the last month or so. A route (almost always a high-interest destination) doesn't get added to the forwarding table correctly.
Not true. The fact that none of my colleagues, the isp team, and I was contacted/involved seems to suggest these routers do not run Cisco IOS. -- Enke ============== [Prev][Next][Index][Thread] Re: C&W Outage, more info, sort of To: nanog@merit.edu Subject: Re: C&W Outage, more info, sort of From: Sean Donelan <SEAN@SDG.DRA.COM> Date: Thu, 3 Jun 1999 14:12:37 -0500 Delivered-To: nanog-outgoing@merit.edu Sender: owner-nanog@merit.edu jmbrown@ihighway.NET (John M. Brown) writes:
It seems that 2 of C&W's core routers in SFO had BGP session problems and they cleared the sessions and everything was ok... Its a "level two issue" at this time and the ticket is still open. L2 folks are looking into the problem to try and figure out what happened.
From our perspective we had no conectivity to NetScape or Yahoo via the C&W link.
Sounds like the cisco bug which has affected several different providers over the last month or so. A route (almost always a high-interest destination) doesn't get added to the forwarding table correctly. Nothing appears wrong at the router, but traffic between individual destinations dies. Unless a customer calls, the provider may not even notice, because it only affects traffic on a particular path through the network. A quick temporary solution is clearing the route table. The router usually builds a fresh route table with the correct forwarding information. And it appears to work until some event tickles the bug later. So far this problem manifest itself only with certain IOS versions using CEF. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation