[slurm-users] Slurmctld 18.08.1 and 18.08.3 segfault

[Bill Broadley]

On 11/13/18 9:39 PM, Kilian Cavalotti wrote:
> Hi Bill,
> There are a couple mentions of the same backtrace on the bugtracker,
> but that was a long time ago (namely
> https://bugs.schedmd.com/show_bug.cgi?id=1557 and
> https://bugs.schedmd.com/show_bug.cgi?id=1660, for Slurm 14.11). Weird
> to see that popping up again in 18.08.

I dug around on bugs.schedmd.com, 1660 is slurmd (not slurmctld crashing), and
1557 seems close, but I never managed to get an assertion failure.

I checked all the other mentions of slurmctld segfault and didn't find anything
particularly close.  I opened ticket 6032.

I was able to clear it by killing jobs and removing nodes from slurm.conf.  I'd
love to be able to track it down to a particular job/node in the future though.