Re: [HTCondor-users] HTCondor-CE || Issue with spool storage size, need SLURM_TMPDIR instead

[Jaime Frey <jfrey@xxxxxxxxxxx>]

You can configure the blahp to run the jobs in another directory. It will create a temporary directory under the alternate location, and copy the input/output files back and forth before/after execution of the job.

Add/modify the following lines to /etc/blah.config:

blah_run_in_workdir=no

blah_wn_temporary_home_dir='$SLURM_TMPDIR'

The single quotes cause the macro expansion to happen on the worker node.

 - Jaime

On Apr 29, 2026, at 8:32âAM, Kamrul Shaker <Kamrul.Shaker@xxxxxxxxxxxx> wrote:

Hello

We are facing one issue with our HTCondor-CE spool storage space.

Its allocated with 200 GB of storage and it cannot be extended further.

Jobs landed on our Slurm cluster utilizes space available in the shared spool directory. Which we don't want.

Will there be any configuration related to this in Blah, where the jobs will dump its payload and output to SLURM_TMPDIR instead of the spool directory?  

Once jobs will be finished, the output will be transferred back to the submit host.

Let me know if you have any questions or concerns. Thanks

Best Regards,

Kamrul Shaker

Cloud and High-Throughput Computing Specialist â IT4 (AESES)
Deanâs Office, Faculty of Science
University of Manitoba
ð Kamrul.Shaker@xxxxxxxxxxxx| ð +1 431 335 3508
ð Winnipeg, MB, Canada
ð www.umanitoba.ca

 

_______________________________________________
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx with a
subject: Unsubscribe

The archives can be found at: https://www-auth.cs.wisc.edu/lists/htcondor-users/