On Tue, 2010-09-21 at 18:46 +0100, Heath Jones wrote: > I've read in a lot of places now that if you have say 192.168.0.0/16 > and then a /24 drops out, that you will end up with 255 /24's in the > table. I'm trying to understand why that is.. (classful implementation > almost?) > My thinking suggests you could have for example one each: /17 + /18 + > /19 + /20 + /21 + /22 + /23 + /24.
As I understand the problem, your interpretaion is correct. Only the worst case de-aggregation of /16 into 256 /24s would cause every even/odd /24 to find a new next-hop. Most cases would be somewhere between x0 (same next-hop) through x8 (your example) and up to x256 (worst case). It would of course take the CPU some time to calculate the new aggregation, and in that period you would potentially have an invalid FIB for more than just the prefix that disappeared. If one of your upstreams (or one of their upstreams) have issues you could see a lot of aggregated prefixes split up like this. And you can't easily know for sure how many in advance. Maybe someone could look at historic instability (from BGPmon?) the past couple of years and model how FIB aggregation would behave? -- Peter _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
