[Bug/_AptManager]: StateSaveLocation file permissions are too restrictive

[NucciTheBoss]

The slurmctld service gets all busted and disgusted when starting up after install because the set permissions on the StateSaveLocation are too restrictive. The current permissions are 0o600 which means that only the user slurm can read/write to the directory:

hpc-libs/lib/charms/hpc_libs/v0/slurm_ops.py

Line 486 in fda80b3

Path("/var/lib/slurm/slurm.state").mkdir(mode=0o600, exist_ok=True)

However, when we try to start the slurmctld service via systemctl, it fails with the error message below:

× slurmctld.service - Slurm controller daemon
     Loaded: loaded (/lib/systemd/system/slurmctld.service; enabled; vendor preset: enabled)
    Drop-In: /etc/systemd/system/slurmctld.service.d
             └─10-slurmctld-nofile.conf
     Active: failed (Result: exit-code) since Wed 2024-11-13 14:44:41 UTC; 2s ago
       Docs: man:slurmctld(8)
    Process: 266530 ExecStart=/usr/sbin/slurmctld -D -s $SLURMCTLD_OPTIONS (code=exited, status=1/FAILURE)
   Main PID: 266530 (code=exited, status=1/FAILURE)
        CPU: 6ms
Nov 13 14:44:41 juju-058453-2 systemd[1]: Started Slurm controller daemon.
Nov 13 14:44:41 juju-058453-2 slurmctld[266530]: slurmctld: fatal: Incorrect permissions on state save loc: /var/lib/slurm/checkp>
Nov 13 14:44:41 juju-058453-2 systemd[1]: slurmctld.service: Main process exited, code=exited, status=1/FAILURE
Nov 13 14:44:41 juju-058453-2 systemd[1]: slurmctld.service: Failed with result 'exit-code'.

It looks that by slightly "loosening" the permissions to be 0o700 where the slurm user can read, write, and execute from StateSaveLocation is the preferred file permission mode. Looking at the slurmd package, it creates a checkpoint directory with the permissions set to 0o755, so it looks like this is what the preferred file permission mode is for StateSaveLocation.