Hello Matt, Thanks. It was the AccountingStorageEnforce parameter indeed. It is working as expected after the parameter was set to "limits,qos" in slurm.conf.
Best, Durai On Fri, Oct 23, 2020 at 6:12 PM Matthew Brown <brown...@vt.edu> wrote: > Yes, I think you need AccountingStorageEnforce to have at least "limits" > set. > > See the AccountingStorageEnforce section here: > https://slurm.schedmd.com/accounting.html > > We use Partition QOS like you described and our "scontrol show config" > shows "AccountingStorageEnforce = associations,limits,qos,safe" > > Best, > > Matt > > > > On Wed, Oct 21, 2020 at 11:22 AM Durai Arasan <arasan.du...@gmail.com> > wrote: > >> Hello, >> >> We recently created a new partition with the following slurm.conf and QOS >> settings: >> >> *cat /etc/slurm/slurm.conf | grep part-long* >> *PartitionName=part-long Nodes=node-1,node-2,node-3 Default=YES, >> AllowAccounts=group1,group2 TRESBillingWeights="gres/gpu=22" MaxNodes=1 >> MaxTime=10-0 QOS=long-10d-jobs State=UP* >> >> >> *sacctmgr show qos format=Name,MaxJobsPU* >> >> >> >> >> * Name MaxJobsPU --------- ---------- normal >> long-10d-+ 2 * >> >> With the above configuration I am able to run 3 jobs simultaneously so >> obviously the partition QOS limit is not being enforced. Do you know why? >> >> I also wanted to ask if AccountingStorageEnforce has any relevance here. >> Because we have not set that in slurm.conf >> >> Thank you, >> Durai >> >>
