RIP not working after client reboot

[BradyPassey]

I'm running a network that relies on RIP, which then I distribute via OSPF through the rest of my network.  If I reboot a client, the route will briefly show up in the routing table and then disappear.  It does this in multiple routers.  The only way to get it to 'stick' and be stable is to clear the routing table.  I can't even clear the specific range, must do a clear ip route *.  For the life of me, I can't figure out why this is happening, anyone have something similar and offer some advice?

[luis_cordova]

Hi @BradyPassey 

The route could disappear because a shorter destination is found. I suspect it could be because there are networks that overlap with others, so I suggest checking the networks and masks, to rule out a false positive.

Regards

[BradyPassey]

I think you may be right, although no routes seem to remain in any routing table.  I have redundant L3 switches that are both passing along the rip via OSPF.  I removed the redistribution on the secondary router and that seems to have resolved it.  Just potentially removes my redundancy should the primary switch fail.

[paul driver]

Hello

Do you have dual mutual redistribution points?

can you post the configuration for your redistribution- 

Are you filtering to prevent route feedback?

[BradyPassey]

Hi Paul,

I'm not sure what filtering to prevent route feedback entails, perhaps that's what I need to investigate.  My config is as follows, where the secondary no longer has the rip redistribution.

router ospf 500
redistribute static subnets
redistribute rip
passive-interface default
no passive-interface Vlan505
network x.x.x.x 0.0.255.255 area 0
network x.x.x.x 0.0.0.255 area 0
network x.x.x.x 0.0.0.255 area 0
maximum-paths 1
default-information originate always

[paul driver]

Hello

Do you have dual redistribution points in you network, meaning is there more than one ingress/egress exit point (rtrs) performing mutual redistribution rip-ospf?

Also
router ospf 500
maximum-paths 1 <-- whats the reason for this 

[BradyPassey]

Yes Paul, I do have multiple in/out doing the distribution.  As far as the maximum paths, that was set by a previous colleague, I believe to avoid (potential) load balancing.

[paul driver]

Hello

okay it is best practice to implement route filtering when you have such a design so not to probably incur route-feedback ( meaning rip into osof into rip

so your rip routes get redistributed into ospf and then those same routes which are now ospf again are redistributed back into rip the same would apply to ospf- rip

Example-

route-map rip-ospf deny 10 

match tag 110

route-map rip-ospf permit 99

set tag 120

route-map ospf-rip deny 10

match tag 120

route-map ospf-rip permit 99

set tag 110

router ospf xx

redistributre rip route-map rip-ospf subnets

route rip

redistribute ospf x route-map ospf-rip metric x

The above would be applied at both redistribution points