OSPF - STATIC redistribution

[bapatsubodh]

Hi,

Setup is simple Router A and Router B are connected to each other by 2MBPS link. On this link OSPF is running.

Router A is connected to other locations

using Channelising E1 serial links. To this locations static routes are provoded using ip route subnet next-hop interface command . ( ip route 10.10.10.0 255.255.255.0 Serial 1/1:1)

Similarlly on Router B many locations are connected using channelised E1 controller. On this router also static routes are used to route the traffic to the locations connected by chanellised E1 links. ( ip route 172.1.1.0 255.255.255.0 serial 0/0:1 ).

Now we have redistributed the static connected routes in the OSPF process so all static routes from both locations are known to each other. So far So good.

Now as a back up to the locations connected over channelised E1 controller card we have used ISDN link. We have configured dialers on both routers. For example location connected to router A will have a backup dialer on ISDN connecting to Router A , ( in case serial link from Chnn. E1 fails ).

Now in order to route the traffic to that location over ISDN connection we have used a secondary static route like

ip route 10.10.10.0 255.255.255.0 192.168.1.2 200. Here in this case 192.168.1.2 is the dialer ip address configured on remote location. and 200 is the admin distance. Here one thing i have found is that when that particular link fails , ISDN takes over and connection is made. I am able to ping to the dialer ip address in our case it is 182.168.1.2 but i am not able to ping to remote routers ethernet address. That is router 10.10.10.1. Then we checked routing table with command

sh ip route 10.10.10.0. To our surrise to showed the route via Router B !

Althought it was connected using static route. This seems to be the problem of the admin. distance we had given in the static route command that is 200. which is more than OSPF. OSPF is running between router A and B. Then we reduced that admin distance to 50. The I got the ping to the router ethernet. But this route is not getting redistributed in the OSPF process. This got confirmed when we have command sh ip route 10.10.10.0 on Router B. It showed subnet not in table.

Hope I have clarified what I want to say.

In a nutshell how do we redistribute seconday static routes in OSPF or normal command redistribute static should work. In our case how do we redistribute the static routes when connected on ISDN.

Thanks

Subodh

Subodh

[sundar.palaniappan]

Subodh,

It would help if you can post the configuration of both routers.

Regards,

Sundar

[ilya.varlashkin]

Do you by any chance have also 'redistribute connected in your config'? I suspect that when you had admin distance of 50 for backup static route it was actually redistributed into ospf, but not installed in the routing table because there was problem with forwarding address. You can see whether a route is redistributed or not by using 'sh ip ospf database' command (redistributed routes will be those at the end denoted as 'external'). Could you please post output of 'sh ip ospf da ext 10.10.10.0' from both routers when backup static route (with admin distance of 200 and 50 are installed)?

The only way I see how with admin distance of 200 a route would point towards router B is router B actually performing redistribution.

Redistribution command doesn't care about admin distance of your routes as long as they're of the type you specified (static in your case).

If backup ISDN link would have been connected to router B, while primary to router A then a bit more elaborate setup would be necessary to ensure that redistributed backup route is less preferred than redistributed primary route. But in your simple case simple 'redistribute static subnet' should be sufficient.

Without seeing configs and 'sh ip rou ' and 'sh ip ospf da ext ' it's rather hard to guess why might have gone wrong. I can only speculate that there is typo in next-hop addresses.

[bapatsubodh]

We have given command redistribute connected subnets. Will it make any difference ?

please suggest.

Thanx.

Subodh

[ilya.varlashkin]

It may. It's quite possible that there is some stale route pointing to problematic network via next-hop address.

Without seeing output of sh route and sh ip ospf database I can't really guess what happens.