On Fri, Nov 08, 2019 at 11:44:24AM +0100, Sacha wrote: > Le 27/10/2019 à 08:54, Claudio Jeker a écrit : > > On Sat, Oct 26, 2019 at 11:33:43PM +0200, Sacha wrote: > >> ------------------------------------------------------------------------ > >> *De :* Stuart Henderson <s...@spacehopper.org> > >> *Objet :* BGPD crash on 6.6 > >> *Date :* samedi 26 octobre 2019 à 21:18 > >> *Pour :* Sacha <sa...@aquilenet.fr> > >> *Cc :* bugs@openbsd.org > >> > >>> On 2019/10/26 15:07, Sacha wrote: > >>>>>> Finaly, it's not a process crash it just a clean shutdown, but it is > >>>>>> not > >>>>>> excepected and we don't know why. > >>>>> "route decision engine terminated; signal 11" is not a clean shutdown, > >>>>> it is SIGSEGV. > >>>>> > >>>>> If you followed Claudio's suggestion of setting sysctl > >>>>> kern.nosuidcoredump=3 > >>>>> and creating /var/crash/bgpd, you should have a bgpd.core file. > >>>>> > >>>> Hi , > >>>> > >>>> Good news: my last try to reproduce the issue didn't crash BGPD. > >>>> > >>>> We have cleaned our bgpd.conf and it fixed the issue. > >>> that is a pity as we would like to get the bug fixed :-( > >> Yes, I didn't had time to try the backups configurations files reproduce > >> the bug. I will be back to the dataceter in 2 weeks to try again an bug > >> hunting. > >> > >> Sacha. > >> > > Would it be possible to share the config with me (working and problematic > > one). Maybe I can figure out where the crash happens when doing a similar > > setup. > > > Hi, > > I tried to reproduce the bgpd daemon bug with no success. > > Maybe OpenBSD 6.6 errata 003 corrected this issue ? > > I left kern.nosuidcoredump=3 and /var/crash/bgpd just in case this bug > is persistent. >
The OpenBSD 6.6 errata 003 is for sure fix one of the crashes that is easy to trigger when sessions go down and nexthops change around the same time. So yes, I'm confident that your issues was solved by this errata. Thanks for rechecking. -- :wq Claudio
