If an OSPF router has two default routes sent to it, via different routers and they are both E1 type routes, does it base its route selection on metric only?
My reason for asking is, I have a router is running two OSPF processes and receives a default route from both via the default-information originate metric-type 1 command from connected routers. The reason for this is that the router participates in a "failover" scenario whereby should the internet connection be lost, all traffic is routed out via another router at a different location. I can explain this in a bit more detail if necessary.
One of these routes has a metric of [110/31] and one has a metric of [110/11].
If i issue the clear ip route * command, the router selects the route with the [110/11] metric and places it into its routing table. This is good. If i shut the interface that connects the router that is sending me that route, it immediately places the [110/31] metric route into the routing table. This is also good.
When i re-enable the shutdown interface, the router does not swap back to using the route with the better metric, unless the clear ip route * command is run. Is this by design? I am struggling to drag my OSPF theory out of the depths of my mind!
More info can follow if necessary,