Hi Michael,

thanks for your explanation. I understand that setting "MaxTRESMinsPerJob=cpu=172800"  will allow (in my case)

-  a job on the full cluster for 6h

-  a job on half of the cluster for 12 hours

But if I do not wont the same user to run at the same time 2 jobs on half of the cluster for 12 hours (and fill in the cluster for long time) how can I limit his running jobs at 172800 minutes*cpu ?

I was looking for something like "MaxTRESMinsPerUser" but do not find such a limitation resource.

Patrick

Le 18/04/2025 à 17:17, Michael Gutteridge a écrit :

Hi

I think you want one of the "MaxTRESMins*" options:

MaxTRESMins=TRES=<minutes>[,TRES=<minutes>,...]
MaxTRESMinsPJ=TRES=<minutes>[,TRES=<minutes>,...]
MaxTRESMinsPerJob=TRES=<minutes>[,TRES=<minutes>,...]
Maximum number of TRES minutes each job is able to use in this association. This is overridden if set directly on a user. Default is the cluster's limit. To clear a previously set value use the modify command with a new value of -1 for each TRES id.

   - sacctmgr(1)

The "MaxCPUs" is a limit on the number of CPUs the association can use.

 -- Michael

On Fri, Apr 18, 2025 at 8:01 AM Patrick Begou via slurm-users <slurm-users@lists.schedmd.com> wrote:

Hi all,

I'm trying to setup a QoS on a small 5 nodes cluster running slurm
24.05.7. My goal is to limit the resources on a (time x number of cores)
strategy to avoid one large job requesting all the resources for too
long time. I've read from https://slurm.schedmd.com/qos.html and some
discussion but my setup is still not working.

I think I need to set these informations:
MaxCPUsPerJob=172800
MaxWallDurationPerJob=48:00:00
Flags=DenyOnLimit,OverPartQOS

for:
12h max for 240 cores => (12*240*60=172800mn)
no job can exceed 2 days
do not accept jobs out of these limits.

What I've done:

1) create the QoS:
sudo sacctmgr add qos workflowlimit \
      MaxWallDurationPerJob=48:00:00 \
      MaxCPUsPerJob=172800 \
      Flags=DenyOnLimit,OverPartQOS


2) Check
sacctmgr show qos Name=workflowlimit format=Name%16,MaxTRES,MaxWall
                Name       MaxTRES     MaxWall
    ---------------- ------------- -----------
       workflowlimit    cpu=172800  2-00:00:00

3) Set the QoS for the account "most" which is the default account for
the users:
sudo sacctmgr modify account name=most set qos=workflowlimit

4) Check
$ sacctmgr show assoc format=account,cluster,user,qos
    Account    Cluster       User                  QOS
---------- ---------- ---------- --------------------
       root     osorno                          normal
       root     osorno       root               normal
       legi     osorno                          normal
       most     osorno                   workflowlimit
       most     osorno      begou        workflowlimit

5) Modifiy slurm.conf with:
     AccountingStorageEnforce=limits,qos
and propagate on the 5 nodes and the front end (done via Ansible)

6) Check
clush -b -w osorno-fe,osorno,osorno-0-[0-4] 'grep
AccountingStorageEnforce /etc/slurm/slurm.conf'
---------------
osorno,osorno-0-[0-4],osorno-fe (7)
---------------
AccountingStorageEnforce=limits,qos

7) restart slurmd on all the compute nodes and slurmctld + slurmdbd on
the management node.

But I can still request 400 cores for 24 hours:
[begou@osorno ~]$ srun -n 400 -t 24:0:0 --pty bash
bash-5.1$ squeue
   JOBID        PARTITION               NAME       USER ST TIME         
START_TIME TIME_LIMIT CPUS NODELIST(REASON)
     147            genoa               bash      begou  R 0:03
2025-04-18T16:52:11 1-00:00:00  400 osorno-0-[0-4]

So I must have missed something ?

My partition (I've only one) in slurm.conf is:
PartitionName=genoa  State=UP Default=YES MaxTime=48:00:00
DefaultTime=24:00:00 Shared=YES OverSubscribe=NO Nodes=osorno-0-[0-4]

Thanks

Patrick


--
slurm-users mailing list -- slurm-users@lists.schedmd.com
To unsubscribe send an email to slurm-users-leave@lists.schedmd.com