On 5/29/26 09:40, Kevin Buckley (Pawsey) via slurm-users wrote:
On Wednesday, May 27th, 2026 at 17:53, Ole Holm Nielsen via slurm-users
<[email protected]> wrote:
The safe way to enable the new purge parameters is to introduce them very
gradually, for example:
PurgeJobScriptAfter=2000days
PurgeJobEnvAfter=2000days
and lower the values little by little over time (restarting slurmdbd each
time).
Do you mean to start off with a value of 2000 days?
Yes, 2000 days purging interval would do no harm and ensure that you don't
overload the slurmdbd server with database operations, so it's good for an
initial test where you check the slurmdbd performance over a couple of
days where daily purging takes place.
I have an extended discussion of Slurm database purging in this Wiki page:
https://wiki.fysik.dtu.dk/Niflheim_system/Slurm_database/#setting-database-purge-parameters
Some sites (although perhaps not mine!) could have completely refreshed
the hardware resourses that Slurm is controlling, within a 5 year period.
The compute node hardware is a separate resource from the Slurm database
server. The age of the Slurm database contents is what matters when it
comes to purging of records.
Would you agree on these considerations?
Best regards,
Ole
--
slurm-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
