[slurm-users] MinCPUsNode in job arrays

Dear SLURM experts, I'm having trouble understanding an issue we have with slurm 17.11.10. In one partition "all", we have some nodes with hypterthreading and some without, leading to 56 and 28 "cores", respectively. In the same partition, we have some nodes with 256GM and some with 128GB R

Re: [slurm-users] problems with slurm and openmpi

I missed this step then to build pmix separately. I thought that the built in pmix inside openpmi could be used by slurm > > On Mar 14, 2019 at 9:32 PM, mailto:gil...@rist.or.jp)> > wrote: > > > > Riccardo, > > > I am a bit confused by your explanation. > > >

Re: [slurm-users] problems with slurm and openmpi

Hello, I installed openmpi from sources and all hte libraries and proper include files where installed correctly in /opt/openmpi/4.0.0 as I prefer it in a directory that can I export via NFS rather than the default /usr/local Anyway slurm's configure still complains and it is not happy ./conf

Re: [slurm-users] problems with slurm and openmpi

Riccardo, I am a bit confused by your explanation. Open MPI does embed PMIx, but only for itself. An other way to put it is you have to install pmix first (package or download from pmix.org) and then build SLURM on top of it. Then you can build Open MPI with the same (external) PMIx or t

Re: [slurm-users] problems with slurm and openmpi

thanks to all. the problem is that slurm's configure is not able to find the pmix includes configure:20846: checking for pmix installation configure:21005: result: configure:21021: WARNING: unable to locate pmix installation regardless of the path I give. and the reason is that configure searche

Re: [slurm-users] [EXTERNAL] slurm-users Digest, Vol 16, Issue 18

Dear Marcus and all peers, Thank you so much for your advice. Lit by your suggestion, after comparing slurm.conf in my cluster to another one which can request one CPU, I noticed that the slurm.conf has not been set properly by manufactory. There were not settings for Socket and CPU (# of socke

[slurm-users] Oversubscribe for reservation

Hello, I was wondering if it is possible to allow oversubscribing for resources in a reservation. For instance, I create a reservation for a user of a specific node. Then I would like to allow that user to do whatever they want, including (say the nodes has 88 cores): srun -n 88 --pty bash a

Re: [slurm-users] How do I impose a limit the memory requested by a job?

Exactly.  The easiest way is just to underreport the amount of memory in slurm.  That way slurm will take care of it natively. We do this here as well even though we have disks in order to make sure the OS has memory left to run. -Paul Edmon- On 3/14/19 8:36 AM, Doug Meyer wrote: We also run

Re: [slurm-users] How do I impose a limit the memory requested by a job?

We also run diskless. In the slurm.conf we round down on memory so slurm does not have the total budget to work with and use a default memory per job value reflecting declared memory/# of threads per node. If users don't declarememory limit we are fine. If they declare more we are fine too. Mostly

Re: [slurm-users] How do I impose a limit the memory requested by a job?

Hello Paul, Thank you for your advice. That all makes sense. We're running diskless compute nodes and so the usable memory is less than the total memory. So I have added a memory check to my job_submit.lua -- see below. I think that all makes sense. Best regards, David -- Check memory/node is va