On 13-10-15 11:26, Bjørn-Helge Mevik wrote:
Restarting the slurmd daemons and/or the slurmctld daemon should in
general not kill jobs.
But if you change things in slurm.conf such that the format of the slurm
state files changes, then restarting slurmctld might result in all jobs
being killed.
This may have been the case; it happened when I enabled GresTypes=gpu.
The man page did mention that "[t]he slurmctld daemon must be restarted
for changes to this parameter to become effective", however I failed to
understand that all running jobs would be terminated by slurmctld.
Indeed the jobs were not terminated by the restart of the slurmd, that
was just required to get slurmctld and slurmd to resume communicating
and immidiately execute the terminations requested by slurmctld.
Robbert
--
Robbert Eggermont Intelligent Systems
r.eggerm...@tudelft.nl Electr.Eng., Mathematics & Comp.Science
+31 15 27 83234 Delft University of Technology
