[slurm-users] Slurm configuration, Weight Parameter

[Renfro, Michael]

We’ve been using that weighting scheme for a year or so, and it works as expected. Not sure how Slurm would react to multiple NodeName=DEFAULT lines like you have, but here’s our node settings and a subset of our partition settings.

In our environment, we’d often have lots of idle cores on GPU nodes, since those jobs tend to be GPU-bound rather than CPU-bound. So in one of our interactive partitions, we let non-GPU jobs take up to 12 cores of a GPU node. Additionally, we have three memory configurations in our main batch partition. We want to bias jobs to running on the smaller-memory nodes by default. And the same principle applies to our GPU partition, where the smaller-memory GPU nodes get jobs before the larger-memory GPU node.

=====

NodeName=gpunode[001-003]  CoresPerSocket=14 RealMemory=382000 Sockets=2 ThreadsPerCore=1 Weight=10011 Gres=gpu:2
NodeName=gpunode004  CoresPerSocket=14 RealMemory=894000 Sockets=2 ThreadsPerCore=1 Weight=10021 Gres=gpu:2
NodeName=node[001-022]  CoresPerSocket=14 RealMemory=62000 Sockets=2 ThreadsPerCore=1 Weight=10201
NodeName=node[023-034]  CoresPerSocket=14 RealMemory=126000 Sockets=2 ThreadsPerCore=1 Weight=10211
NodeName=node[035-040]  CoresPerSocket=14 RealMemory=254000 Sockets=2 ThreadsPerCore=1 Weight=10221

PartitionName=any-interactive Default=NO MinNodes=1 MaxNodes=4 MaxTime=02:00:00 AllowGroups=ALL PriorityJobFactor=3 PriorityTier=1 DisableRootJobs=NO RootOnly=NO Hidden=NO Shared=NO GraceTime=0 PreemptMode=OFF ReqResv=NO DefMemPerCPU=2000 AllowAccounts=ALL AllowQos=ALL LLN=NO MaxCPUsPerNode=12 ExclusiveUser=NO OverSubscribe=NO OverTimeLimit=0 State=UP Nodes=node[001-040],gpunode[001-004]

PartitionName=batch Default=YES MinNodes=1 MaxNodes=40 DefaultTime=1-00:00:00 MaxTime=30-00:00:00 AllowGroups=ALL PriorityJobFactor=1 PriorityTier=1 DisableRootJobs=NO RootOnly=NO Hidden=NO Shared=NO GraceTime=0 PreemptMode=OFF ReqResv=NO DefMemPerCPU=2000 AllowAccounts=ALL AllowQos=ALL LLN=NO ExclusiveUser=NO OverSubscribe=NO OverTimeLimit=0 State=UP Nodes=node[001-040]

PartitionName=gpu Default=NO MinNodes=1 DefaultTime=1-00:00:00 MaxTime=30-00:00:00 AllowGroups=ALL PriorityJobFactor=1 PriorityTier=1 DisableRootJobs=NO RootOnly=NO Hidden=NO Shared=NO GraceTime=0 PreemptMode=OFF ReqResv=NO DefMemPerCPU=2000 AllowAccounts=ALL AllowQos=ALL LLN=NO MaxCPUsPerNode=16 QoS=gpu ExclusiveUser=NO OverSubscribe=NO OverTimeLimit=0 State=UP Nodes=gpunode[001-004]

=====

> On Nov 29, 2019, at 8:09 AM, Sistemas NLHPC <sistemas at nlhpc.cl> wrote:
> 
> External Email Warning
> This email originated from outside the university. Please use caution when opening attachments, clicking links, or responding to requests.
> Hi All,
> 
> Thanks all for your posts 
> 
> Reading the documentation of Slurm and other sites like Niflheim https://wiki.fysik.dtu.dk/niflheim/Slurm_configuration#node-weight (Ole Holm Nielsen) the parameter "Weight" is to assign a value to the nodes, with this you can have priority in the nodes. But I have not obtained positive results.
> 
> Thanks in advance
> 
> Regards
> 
> El sáb., 23 nov. 2019 a las 14:18, Chris Samuel (<chris at csamuel.org>) escribió:
> On 23/11/19 9:14 am, Chris Samuel wrote:
> 
> > My gut instinct (and I've never tried this) is to make the 3GB nodes be 
> > in a separate partition that is guarded by AllowQos=3GB and have a QOS 
> > called "3GB" that uses MinTRESPerJob to require jobs to ask for more 
> > than 2GB of RAM to be allowed into the QOS.
> 
> Of course there's nothing to stop a user requesting more memory than 
> they need to get access to these nodes, but that's a social issue not a 
> technical one. :-)
> 
> -- 
>   Chris Samuel  :  http://www.csamuel.org/  :  Berkeley, CA, USA
>