[c-nsp] General BCP for ibgp maximum-paths to 0.0.0.0

Jason Lixfeld jason at lixfeld.ca
Thu Oct 18 16:37:37 EDT 2012


What's the general consensus for how a low-memory BGP speaker, multi-homed to two full-table-sucking iBGP speakers should receive and use a default route from both full-table-sucking iBGP speakers?

The general idea in my head is pretty straight forward - big-ass routers originate a default towards low-mem router in the VRF where the default is required, low-mem router enables ibgp maximum-paths 2 in the same VRF and Bob's your uncle.  Is that still pretty much the defacto standard?

Any arguments for/against ip route 0 null0 ; router bgp x ; network 0 vs ip route 0 null 0; router bgp x ; redistribute static vs some other method?

Thanks in advance..





More information about the cisco-nsp mailing list