>From what I could dig up, yes, 1 per second, rate limited on a per-interface basis.
>From a CEF point of view, when we do the lookup and realize we have to do a redirect we punt to process switching. From the 6k perspective this would be a specific punt to software that would be matched against the MLS rate limiter. At that point the ICMP code looks at the input interface and the last time a redirect was generated on that interface to determine if a new redirect should be generated. -Pete On Tue, Feb 12, 2013 at 4:46 AM, Phil Mayers <[email protected]>wrote: > On 02/12/2013 02:34 AM, Pete Lumbis wrote: > >> I'm pretty sure this is statically defined within IOS for all redirects >> on an given interface. >> > > So it's 1 redirect per second per input-interface, not configurable? > > If it's an IOS-wide behaviour, that implies it's not platform specific, > which would make sense, and ties with my original understanding of the way > redirects are handled on this platform (all packets to sup, sup forwards in > software). > _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
