We actually implemented this in IBM's Nways products in the early/mid nineties and found it wasn't worth the trouble. It had no measurable effect on network performance and the protocols necessary to terminate the broadcast were messy and error prone.
Sent from my iPhone > -----Original Message----- > From: [email protected] [mailto:[email protected]] On Behalf Of > András Császár > Sent: Tuesday, April 05, 2011 6:09 AM > To: Bhatia, Manav (Manav); [email protected] > Cc: [email protected] > Subject: Re: [OSPF] draft-kini-ospf-fast-notification-01 > > > I believe the idea is quite simple - > > Flooding a changed LSA in HW (data plane) is faster than > > doing it in SW (not matter how optimized the latter is) or > > your control plane. > > That's the main purpose, yes. > > You could combine this with something else: OSPF could pre-calc/pre- > download potential new LSAs to the dataplane anticipating local > failures. > > Whenever a local failure really happens, dataplane notices it by LoS or > BFD, then can immediately disseminate the changed LSA using FN. So, in > addition to the per-hop CP processing of LSAs, you could also spare > some time with originating an LSA: > - failure trigger reaching/processed by CP > - OSPF originating a new LSA > - OSPF sending the LSU packet down to DP > > Cheers, > András > _______________________________________________ > OSPF mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/ospf _______________________________________________ OSPF mailing list [email protected] https://www.ietf.org/mailman/listinfo/ospf
