BGP OSPF distribution

Unanswered Question
Jun 17th, 2007
User Badges:

We have a fialover situation that I am trying to test. Supposedly it has worked in the past, but does not seem to be working at the moment.


Under normal operation at the HQ site, the Internet router is receiving a default route from its peer MCI router via BGP. This route is redistributed into OSPF and transmitted to the HQ PIX and then to the rest of the internal network. In the event the Internet router or Internet circuit fails, the HQ PIX will no longer receive and no longer propagate a default route to internal devices.


Also, under normal operation, the DR Internet router is receiving a default route from its peer MCI router via BGP. This route is redistributed into OSPF and transmitted to the DR PIX and then to the rest of the network devices at DR. This route is then redistributed into BGP on the DR 3725 and is transmitted to the 7206 in the HQ site. The BGP configurations on the 3725 router at DR include a ?distance? command which offsets the route metric and makes it less favorable to other devices than the route which is being redistributed into the HQ network via the Internet router in HQ. Therefore, the default route which is learned via the DR Internet router is only used when the HQ Internet router is down, or otherwise stops advertising its default route.


The above description is from the consultants that built the network.


Supposedly this has been tested and works, but:


I shut down the serial Interface on the HQ Internet router and none of the HQ devices switched their default route to the DR Internet.


All of the HQ devices still were pointing out toward the HQ Internet, event though the link was down.


I noticed the DR devices were pointing out the the DR Internet, but the HQ devices continued to point to the HQ Internet.


I also noticed the same distance command on the HQ 7206 router as the DR 3725 described above.


Would the HQ devices continue to hold the default route even though they were not getting updates from the HQ Internet router?


Or should the routes have died and the devices had no defualt route if there was route distribution problem between the two seperate locations?


Both locations are using their own Internet as the default route for all their respective network devices.

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
cablexpress Thu, 11/08/2007 - 12:15
User Badges:

Did you ever get a resolution to this issue. I'm having the same problem.

wilson_1234_2 Fri, 11/09/2007 - 13:42
User Badges:

I did get this problem resolved, which turned out to be several problems rolled into one.


One of the problems was that one of the core switches had a static default route pointing to the PIX firewall, and also had "default information-originate" configured in the OSPF config.


Every neigbor to the switch was pointing to it for their default route.


When the gateway on this switch was supposed to flip to the DR site, it was locked by the static route, which it also advertised to it's neighbors.


When I removed the static route and "default information-originate", things started to fall into place.

Actions

This Discussion