More broadly, it is delay in general that is a problem for all the IGPs I have tested to date. Whether it be processing time of routes (cpu bound), or network bound (slow wifi multicast), really bad things start to happen once the window for processing events is exceeded. I wrote this readme in exasperation years and years ago. https://github.com/dtaht/rtod#rtod-routing-tables-of-death The same holds true for bgp, as well. I like that babel has a hitherotofore unused facility for dynamically managing the announcement intervals, and hope one day more demons of all sorts, with implicit processing deadlines, do explicit things to shed load intelligently, rather than fail randomly. We have long had facilities in linux at least (fdtimers) to better monitor exceeding deadlines. None of what I said has any bearing on the comments you are discussing! -- last-call mailing list last-call@xxxxxxxx https://www.ietf.org/mailman/listinfo/last-call
