Hi Mathieu,
maybe as an alternative: would it work for you to bind mount /tmp (or
${TMPDIR} more geenral) into the job directory itself? (requiring 8.6.1
and vanilla or docker universes [1]).
With
MOUNT_UNDER_SCRATCH
and maybe setting an upper limit on the job disk quota, one could keep
and organize all the job tmp output at one place.
Cheers,
Thomas
[1]
http://research.cs.wisc.edu/htcondor/manual/v8.6/10_3Stable_Release.html
On 2018-04-04 09:56, Mathieu Bahin wrote:
> Hi,
>
> Is there a way to limit the number of jobs from one batch per node?
>
> I'm running jobs that use the /tmp/ space and I don't want too much of
> them on a same node at the same time.
> We have a lot of nodes so it's not a problem to spread them but I would
> like to specify this constraint to be sure it's effective.
>
> Cheers,
> Mathieu
>
Attachment:
signature.asc
Description: OpenPGP digital signature