Our network runs OSPF internally. The topology is basically a ring from our distribution routers to core routers in the data center. (Were a small SP). Distribution connects to either another distribution via L3 links, or to a mix of distribution and core, depending on which link in the ring we’re talking about.
We’ve recently completed building our redundant WAN backbone ring and are doing some fail over testing. Originally everything worked as expected. I could disconnect any link throughout the ring and OSPF would drop routes from dead neighbors and inject backup routes, all was well.
Tonight I was testing the fail over again by removing a link in the ring and the routing table is keeping its default route via the broken link. The OSPF neighbor on the broken is dead, so the default route should withdraw and fail over to the alternate path, right?
I tried clearing the process to see if it was hung and that didn’t effect it. I also noticed this happens on more than one router. So maybe I’m just missing something here?
It’s late and I’m sure I’m forgetting information, so my apologies in advance. Btw we’re a 100% Cisco shop in case you think it’s a vendor specific bug or something.
No comments:
Post a Comment