[slurm-users] slurmctld/slurmdbd filesystem/usermap requirements
pbrunk at uga.edu
Thu Feb 10 13:11:17 UTC 2022
slurmctld runs as an unprivileged user ('slurm' by default) who probably doesn't have read access to the user's job scripts. 'sbatch' submits the scripts via network to slurmctld, who stores them in the slurm.conf 'StateSaveLocation', and sends them to slurmds at dispatch time, who store them on each node in the slurm.conf 'SlurmdSpoolDir', as Steffen noted.
All this to say that the slurmctld host doesn't need to see the users' home dirs and/or job script dirs.
Paul Brunk, system administrator
Georgia Advanced Resource Computing Center
Enterprise IT Svcs, the University of Georgia
On 2/10/22, 6:26 AM, "slurm-users" <slurm-users-bounces at lists.schedmd.com> wrote:
[EXTERNAL SENDER - PROCEED CAUTIOUSLY]
On Thu, 2022-02-10 at 11:59:58 +0100, Diego Zuccato wrote:
> Hello all.
> Does slurmctld (or slurmdbd) need to access the same filesystems used on
> submit nodes? Or they just receive the needed information in the request?
> Does slurmctld need read access to /home/userA/myjob.sh or does it receive
> the job script as a "blob" or as a path? Does it even need to know userA's
> GID or will it simply use 'userA' to lookup associations in dbd?
Looking into one of our Slurm installations (pre-installed):
- file systems must not be shared with the machine running the dbd/ctld,
the job script is copied into a subdirectory of /var/spool/slurm on
the exec node (but I can't say for sure who does this)
- UIDs are reported by the ctld so they "should" be identical for the
controller and submit machines (and the execute ones, of course) -
I never saw GIDs mentioned though.
Does this help?
Steffen Grunewald, Cluster Administrator
Max Planck Institute for Gravitational Physics (Albert Einstein Institute)
Am Mühlenberg 1 * D-14476 Potsdam-Golm * Germany
Fon: +49-331-567 7274
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the slurm-users