So I'm working with a company who have two BGP peerings terminating on two edge switches. Long story short because of black hole routes we need to export both the default route and a /x into the network. That way if one uplink goes down, the blackhole doesn't eat the traffic, it uses ospf. This theory worked until i went to implement it. It worked perfectly until we disabled/reenabled the ports. At which point rather than the original bgp route coming back and the "secondary" ospf link not being used, the bgp route never became active.Interestingly the default route did though.
Am I being a fool or is there something odd I need to do?
BTW, I used a policy to export the routes to ospf.. not sure if that makes a difference.