When I issue a srun or sbatch command against a suspended partition, the
job is not allocated to any nodes so nothing gets powered on.  If I
manually execute the slurm_resume file or if I set the state with scontrol
to power on the slurm_resume script executes as expected.

Any thoughts on why suspended nodes do not get allocated to a job via srun
or sbatch?  I'm assuming this is why slurm_resume isn't running.

Matt

Reply via email to