Sunday, May 9, 2010

troubleshooting

Here's a problem I ran into. I set up a new environment with a different layer 3 infrastructure. And everything seemed to be working. Until I got reports that some people couldn't get to one network in the new environment.

So, traffic going to the range 10.2.84.0/24 was working great for some people, but for others, it was not -- TCP connections would connect and then fail after a few seconds.

I spent a bunch of time looking at switch ports, and spanning tree, looking to see where the blocking ports were and making sure that there's no loops. So I gave up on that line of inquiry and started tracing back at layer 3.

Sure enough, there were two routes for 10.2.184.0/24 in one of my core routers - one of which pointed to the right place, and one of which pointed to the wrong place.

I deleted the wrong route, and things worked again.

2 comments:

  1. Out of curiosity, what generated the incorrect route?

    ReplyDelete
  2. Me!
    Regis sits in the corner with the dunce hat on

    More specifically, the new environment was actually a data center move that included some topology changes. And one of the topology changes was a change in that route -- I'd remembered to add the new route, but forgot to delete the old one.

    ReplyDelete

Followers

About Me

My photo
Regis has worked as a network engineer since 1994 for small companies and for large companies.