David: Created page with "We had a situation where a user job kept on resubmitting the scheduler, with the same JOBID. So everytime the job failed, the job info in slurm was overwrote. We found out..."

2015-07-08T20:26:45Z

Created page with "We had a situation where a user job kept on resubmitting the scheduler, with the same JOBID. So everytime the job failed, the job info in slurm was overwrote. We found out..."

New page

We had a situation where a user job kept on resubmitting the scheduler, with the same JOBID. So everytime the job failed, the job info in slurm was overwrote.

We found out that Slurm is configured to automatically requeue jobs which were aborted due to node failures.

If this is an unwanted behaviour you can prevent automatic requeuing with the following option in your job script:

<syntaxhighlight>
#SBATCH --no-requeue
</syntaxhighlight>

Slurm: Prevent jobs from being resubmitted - Revision history

David: Created page with "We had a situation where a user job kept on resubmitting the scheduler, with the same JOBID. So everytime the job failed, the job info in slurm was overwrote. We found out..."