> Or, does the traffic hit the MSFC because the CEF adjacency is the same > interface for all prefixes, i.e. router-on-a-stick? > > I've read references to both as being the reason one would see high CPU on > an MSFC along with SVI input drops, just trying to understand the hardware > better. It seems the magic number is around 40-50 Mbps which resulted in > 75% CPU usage and input drops. That just seems low for such powerful > hardware...
you either have a feature enabled which isn't implemented in hardware (hence h/w has to punt to software to process), or you've hit some kind of bug that is causing your particular traffic to be sent to software for processing. beyond mentioning the hardware you have, you haven't mentioned what IOS release, nor have you posted any configuration snippet from your SVI, so it's a little difficult to provide any meaningful diagnosis as to which of the above it may be. as a suggestion, if you have a support contract, open a TAC case to troubleshoot with the TAC. alternately, have a look through release notes for newer IOS releases and see if you can find a 'closed bug' that matches the behavior you have seen. alternately, search through the bug database to see what you can find. I don't think "40Mbps" is a magical number. it just represents some pkts/sec level. as a general statement, you rarely want to be hitting software switching, less so process-level switching within software switching! cheers, lincoln. _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
