On 04/05/11 09:29, Vitkovsky, Adam wrote:
Why do you need to advertise multicast routes over BGP?
It's for RPF to work.
I'm not sure now but wouldn't the rpf failover to unicast safi if the route
can't be find in the m-cast safi?
Not necessarily.
If you have *no* routes in your inet.2 (i.e. multicast rpf) table, then
it might use inet.0 (i.e. unicast routing table) - I can't honestly
remember JunOS behaviour in this case (I assume so - because we have
MVPN working on JunOS without inet.2)
But if you have *any* routes in inet.2, you probably need *all* routes.
That's the point of having a separate multicast RPF table - you might
want the RPF topology to be different to the unicast/forwarding
topology, so you can't just assume that unicast is OK to fall back to.
_______________________________________________
cisco-nsp mailing list [email protected]
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
