From: Marko Milivojevic (markom@markom.info)
Date: Wed Jul 02 2008 - 20:43:41 ART
A wild shot in the dark here... On that router that is experiencing
problems, are you perhaps running this OSPF in a VRF (VRF-lite)?
To do more precise troubleshooting, interface and router process
configurations would be needed. Show ip route, show ip ospf nei, show
ip ospf int, etc. would certainly help, too.
On Wed, Jul 2, 2008 at 8:38 PM, Michael Whittle <mgwhittle@gmail.com> wrote:
> Hi all,
>
> I wonder if you can help me.
>
> I have an OSPF problem and a little stuck. I have an MPLS network with
> multiple spoke VPN's. I have a Cisco ASA5510 connected to a PE with a trunk.
> Each VPN connects on it's own VLAN. Each connection between the ASA and the
> PE is running OSPF area 0. The ASA only advertises a default route and a
> couple of other routes to each MPLS VPN spoke. The spokes only advertise
> their BGP routes into OSPF using a route-map to filter. Each connection to
> the ASA is configured the same way and it's working perfectly except for one
> of them.
>
> For some reason the ASA is seeing routes flapping on one of them when the
> configurations are identical and everything is shared. It seems to go in
> about a 10 second interval. All routes received, then all routes lost. At
> first I thought it must be a routing loop but usually when this happens you
> would see the routes changing rather than flapping. OSPF is stable and the
> neighbor is up. I'm not seeing anything nasty in the logs. If I add static
> routes on the ASA towards the VPN then it's stable so that rules out the BGP
> routes having issues. My question is what could cause this sort of behavior
> where the routes appear and disappear in cycles like this? Is there anything
> I can check or any useful debugs I can try? I tried "debug ip ospf adj",
> "debug ip ospf spf" and "debug ip ospf events" and none of them show
> anything useful.
>
> I would have thought if there was a problem with OSPF the adjacency would be
> flapping as well. If it was a routing loop surely I wouldn't see the route
> leave all together just the next-hop be updated by other device. So I'm just
> trying to work out what else it could be. I would really appreciate some
> tips on what I could try.
>
> Thanks in advance.
This archive was generated by hypermail 2.1.4 : Mon Aug 04 2008 - 06:11:53 ART