Hello,
Simple question.
When a failure occurs on backbone, OSPF does its thing and stops forwarding traffic to failed gateway creating a new route to a different gateway. If MPLS is not in action, everything would work excellent!
However MPLS is unable to keep up with OSPF in case of failovers. Its not removing the old forwarding base entry which maps destination networks to old failed gateway, it stays there forever.
MPLS process is supposed to delete all forwarding base entries that are associated with a non-existent gateway in routing table once a topology change occurs.
Has someone had the same issue? Reproduced it in lab? This seems to be a bug. MPLS is clearly acting as a blackhole.

Thanks

If you can reproduce it in the lab, send us supouts and topology of lab network.