OSPF/RIP Redistribution

[davem1]

Good afternoon,

I've got a GNS3 project where I've got RT01/RT02 forming an OSPF adjacency with HOUSTON.  RT01/RT02 then connect to CPE01/CPE02 where the routes are redistributing via RIPv2.

RT01 is receiving the RIP routes via CPE01/CPE02; there is no issue there. If I shut down the F0/1.415 interface the RIP routes are then populated on router 2 via CPE01/CPE02.  My issue is this, when i re-enable F0/1.415 on RT01 the RIP routes do not re-populate in the routing table; the routes are still in the routing table on RT02. In order to have the routes go back to RT01 I have to shut down the F0/1.415 interface on RT02.

My question is if it is possible for the routes to re-advertise back to RT01 automatically.

Thanks.

[Julio E. Moisa]

Hi

Yes, you can, basically you need to understand the metrics for each routing protocol, when you redistribute OSPF into RIP you need to configure a metric, the same case when you redistribute RIP into OSPF, the lowest metric will always preferred as the best path. 

RIP is not manageable like EIGRP or OSPF or other protocols but you can redistribute the OSPF routes into RIP with different hops in order to prefer a path over other. 

You could have:

Router 1

router ospf 410 vrf CUST
redistribute rip metric-type 1 metric 100 subnets

!
router rip
address-family ipv4 vrf CUST
redistribute static
redistribute ospf 410 metric 1

Router 2

router ospf 410 vrf CUST
redistribute rip metric-type 1 metric 200 subnets

!
router rip
address-family ipv4 vrf CUST
redistribute static
redistribute ospf 410 metric 3

[davem1]

I applied those changes but they made no difference.

I can receive the RIP routes from CPE01/02 (it's usually one or the other), but the problem is that the RIP routes do not automatically prefer RT01. If I shut down F0/1.700 on RT01 the RIP routes proceed to RT02, but when I re-enable the interface the routes remain on RT02 and do not automatically go back through RT01. The only way to correct it is to shut down F0/1.700 on RT02 so that the routes can revert back to RT01.

I've re-attached the router configurations with what they are currently running. Also, I had an overlapping subnet on the CPE routers which I have also corrected.

Thanks for help.

[Julio E. Moisa]

Hola David,

Ok please let me lab it and find the problem. 

[davem1]

Thanks, Julio. I think I found the issue. While debugging RIP on RT01 I did notice that we were receiving the routes but they were not showing in the routing table. The reason being that they were already learned via OSPF with the higher AD. I implemented an inbound distribute-list in the OSPF instances (acl_DENY_RIP_ROUTES) on both routers and it seems to be working.

I would like to know if you came to the same conclusion or found another way in which the issue was being caused.

Regards,