On May 12, 2010, at 6:21 AM, Phil Mayers wrote: > On 05/12/2010 10:55 AM, Youssef Bengelloun-Zahr wrote: > >> We have a bunch 6509s acting as core routers and a bunch of 7204VXRs >> (NPE-400 / NPE-G1) acting as LNS border routers. >> >> Problem Is : I am concerned about performance issues. Is uRPF a big consumer >> of CPU / Memory ? > > On 6500, I believe the older sup2 has half the routing table capacity with > uRPF enabled, but it's otherwise done in hardware. > > 6500/sup720 uRPF is "free". There's no CPU/memory load. > > I don't know about the other platforms but I would be surprised if uRPF > significantly affects their forwarding performance.
FYI: In sup720, if you already have interfaces in loose mode and toggle one to strict any interface with uRPF enabled goes to strict. re: SUP2, you are correct, tcam is halved in size from 256k to 128k when uRPF is enabled, if you are running bgp this can trigger software switching of packets and high cpu utilization. - Jared _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
