Hello,
I'm looking into why some jobs are getting cancelled/requeued on my
cluster. The default hypothesis is that it is priority (QOS) preemption,
which was recently turned on. But it seems to be happening way more than it
should based on how many jobs are actually being submitted to a
preemption
Have you considered the --wrap option of sbatch?
*--wrap=*
*Sbatch will wrap the specified command string in a simple "sh" shell
script, and submit that script to the slurm controller. When --wrap is
used, a script name and arguments may not be specified on the command line;
instead the sbatch-gen
Hello,
I'm looking for some feedback on a scheme I've come up with to configure
priority tiers on our cluster which is used by a few dozen groups.
The desired behavior is that users can submit jobs to one of three priority
tiers:
- Tier 1: Lowest priority; any user can use as many resources a
, as there is no advantage to
> using it over tier3.
>
>
> On 04/08//2017 17:30, Evan Remington wrote:
>
> Hello,
>
> I'm looking for some feedback on a scheme I've come up with to configure
> priority tiers on our cluster which is used by a few dozen groups.
>