We have BGP on a 1072 with 2 providers, 1 full routing table and 1 that just ~30 routes… then we static route internally.
What we’ve noticed is that our SNMP monitoring of the vlan facing the provider seems to have some major issues like its timing out or something when we monitor it. Whats even more odd is it seems like the more traffic an interface has the more the snmp for that interface times out, we’re polling every 5 seconds for live stats.
My question is this i wanted to attempt to filter out the full table to only be the default route, will doing accept 0.0.0.0 then drop all as inbound filter drop our CPU load or will it still be getting hit due to the provider still sending the routes to us and us filtering them out? Figured it would be good test to have only default route instead of full table and see if its the BGP updates giving us issues.