[c-nsp] IOS XR BGP

[Nick Hilliard]

On 24/11/2011 16:23, Nick Ryce wrote:
> I understand this but there is sometimes hundreds of more specifics in
> IGP so there would not really be any flapping as such unless some core
> links horrifically died.
> 
> So would redist ospf into BGP then use aggregate address work then?

So for each routing flap, you want your router to calculate updates for
both OSPF and BGP?  Wouldn't recommend this either. :-)

Have you considered changing your architecture to use ibgp for carrying
your interior prefixes instead of ospf?  This works quite well in most
situations, and would give you the advantage that you don't have to start
doing ugly stuff like redistributing ospf into bgp (or worse, vice-versa).
 All you really need in your IGP are the loopbacks for each router.

Nick