In message <9409302214.AA19663@ravn.runit.sintef.no>, Havard Eidnes writes:
Hi,
we have had some problems recently apparently caused by route flapping, and this inspired me to create a setup where I can log the "route flaps of the world" and then do some perl programming to analyze the debug logs. After about a day of programming I now have a handfull of scripts which are able to post-process the output of "debug ip bgp updates" on a Cisco router.
Havard, Nice work. I've done something similar. Reports are sent daily (if we don't run out of disk space processing them) to any AS that wants them. A archive (often a few weeks out of date) is kept at: ftp://ftp.ans.net/pub/info/routing-stats/daily-reports Try the README file first. The idea behind these (quite detailed) reports is to identify which nets or downstream AS is unstable and as a heads up to check into things further. A few well connected AS get more than their share of reports for stuff they have no control over, but I plan to correct that by allowing more per report destination configuration. Curtis