Skip to main content

Quality of Service (QOS)

Quality of Service (QoS) is used to limite the characteristics that a job can have.

Default QOS per partition

For every partition (fct, gpu, etc), there is a Quality of Service with different defined parameters like MaxJobs, MaxSubmitJobs. This parameters act on all the jobs submitted by all user's on that partition. To see the the default QOS for a specific partition run the below scontrol command:

fct partition

[jpina@cirrus02 ~]$ scontrol show partition fct
PartitionName=fct
   AllowGroups=ALL AllowAccounts=fct1,cpca27902020,cpca59032020,cpca3952082021,cpca3949842021,cpca4021052021,cpca3969692021,cpca097952021,cpca4076702021,biosim,cpca4209172021,cpca4081432021,cpca4011972021,dsaipa00832020,cpca098302021,cpca097822021,cpca097522021,cpca096232021,cpca097312021,cpca097642021,cpca230372022,cpca262792022,cpca158802022,cpca158802023,cpca56132020,cpca156102022,cpca158542022,cpca280462022,cpca096232021,cpca159122022 DenyQos=normal,low,medium,high
   AllocNodes=ALL Default=NO QoS=N/A
   DefaultTime=NONE DisableRootJobs=NO ExclusiveUser=NO GraceTime=0 Hidden=NO
   MaxNodes=UNLIMITED MaxTime=4-00:00:00 MinNodes=0 LLN=NO MaxCPUsPerNode=90
   Nodes=hpc06[0-3]
   PriorityJobFactor=1 PriorityTier=1 RootOnly=NO ReqResv=NO OverSubscribe=NO
   OverTimeLimit=NONE PreemptMode=OFF
   State=UP TotalCPUs=384 TotalNodes=4 SelectTypeParameters=NONE
   JobDefaults=(null)
   DefMemPerCPU=5000 MaxMemPerNode=UNLIMITED

gpu partition

[jpina@cirrus02 ~]$ scontrol show partition gpu
PartitionName=gpu
   AllowGroups=ALL AllowAccounts=aeon,biomeng,biosim,cbmr,ccmar,cedis,centec,cerberos,chlab,ciceco,ciimar,cncb,comics,cosmos,csys,dei,eeisel,eworm,dosimetry,fcneuro,fctunlrequinte,fculbioisi,fculce3c,fculdi,fculfisica,fculgfm,fculibeb,feno,hpc,ibb,ibet,ihmt,inl,inov,ipfn,insa,isctesociologia,ispa,istcftp,lapmet,lasige,lnec,lnecprd,localmaxs,mcfeup,neuro,nlx,nps,scipion,seatox,solarb,spac,t3atlas,t3cms,ua,uaberta,uait,uaquimica,ubim,uc,uccibit,uedi,ulcefisa,ulibeb,ulusofona,um,unlims,unlitqb,xtal,yeastgenomics,cpca27902020,cpca59032020,fct1,cpca4209172021,cpca262792022,cpca56132020,cpca158542022,cpca280462022,cpca097822021,cpca159122022 AllowQos=normal,gpu3952082021,gpu4021052021,gpu4209172021,gpu262792022,gpu158802022,gpu158002022,gpu158542022,gpu280462022,fct1,gpu159122022,gpu097822021
   AllocNodes=ALL Default=NO QoS=N/A
   DefaultTime=NONE DisableRootJobs=NO ExclusiveUser=NO GraceTime=0 Hidden=NO
   MaxNodes=1 MaxTime=4-00:00:00 MinNodes=0 LLN=NO MaxCPUsPerNode=2
   Nodes=hpc06[0-3]
   PriorityJobFactor=1000 PriorityTier=1000 RootOnly=NO ReqResv=NO OverSubscribe=NO
   OverTimeLimit=NONE PreemptMode=OFF
   State=UP TotalCPUs=384 TotalNodes=4 SelectTypeParameters=NONE
   JobDefaults=(null)
   DefMemPerCPU=5000 MaxMemPerNode=UNLIMITED

hpc partition

[jpina@cirrus02 ~]$ scontrol show partition hpc
PartitionName=hpc
   AllowGroups=ALL AllowAccounts=aeon,biomeng,biosim,cbmr,ccmar,cedis,centec,cerberos,chlab,ciceco,ciimar,cncb,comics,cosmos,csys,dei,dosimetry,eeisel,eworm,fcneuro,fctunlrequinte,fculbioisi,fculce3c,fculdi,fculfisica,fculgfm,fculibeb,feno,hpc,ibb,ibet,ihmt,inl,inov,ipfn,insa,isctesociologia,ispa,istcftp,lapmet,lasige,lnec,lnecprd,localmaxs,mcfeup,neuro,nlx,nps,scipion,seatox,solarb,spac,t3atlas,t3cms,ua,uaberta,uait,uaquimica,ubim,uc,uccibit,uedi,ulcefisa,ulibeb,ulusofona,um,unlims,unlitqb,xtal,yeastgenomics AllowQos=normal
   AllocNodes=ALL Default=YES QoS=N/A
   DefaultTime=NONE DisableRootJobs=NO ExclusiveUser=NO GraceTime=0 Hidden=NO
   MaxNodes=UNLIMITED MaxTime=4-00:00:00 MinNodes=0 LLN=NO MaxCPUsPerNode=UNLIMITED
   Nodes=hpc04[6-8]
   PriorityJobFactor=1 PriorityTier=1 RootOnly=NO ReqResv=NO OverSubscribe=NO
   OverTimeLimit=NONE PreemptMode=OFF
   State=UP TotalCPUs=192 TotalNodes=3 SelectTypeParameters=NONE
   JobDefaults=(null)
   DefMemPerCPU=8000 MaxMemPerNode=UNLIMITED

User QOS

By default there are no user QOS atributed. All user QOS atributed by INCD team are uniq to specific users. The following comand show how to check the QOS you belong to:


[martinsj@cirrus02 ~]$ sacctmgr show user fmartins withassoc -p
User|Def Acct|Admin|Cluster|Account|Partition|Share|Priority|MaxJobs|MaxNodes|MaxCPUs|MaxSubmit|MaxWall|MaxCPUMins|QOS|Def QOS|
fmartins|biosim|None|production|cpca097522021||1||||||||cpca097522021,gpu097522021,normal||
fmartins|biosim|None|production|cpca097822021||1||||||||cpca097822021,gpu097822021,normal||
fmartins|biosim|None|production|biosim||1||||||||cpca71402020,incdbiosim21,normal||