I have an OSPF network where my BGP routers have non-iBGP routers between them. Because of physical path diversity, I will be using IPIP tunnels to "backhaul" traffic between the BGP routers (for best exit.)
With the next-hop being the loopback address, I fear the routers may try sending traffic to the middle ospf routers instead of the tunnel. Is there a way to modify the forwarding table to prevent sending traffic whose next-hop is the other BGP router to the OSPF routers whose routing tables are not complete?
I would really like to use the loopback as next-hop in the routing table; so while using the tunnel's endpoint as the next-hop address would work, it's my "nothing else works, so I have to" backup plan.
No comments:
Post a Comment