[slurm-users] Sharing a node with non-gres and gres jobs

[Christopher Samuel]

On 3/20/19 9:09 AM, Peter Steinbach wrote:

> Interesting enough, if I add Cores=0-1 and Cores=2-3 to the gres.conf 
> file, everything stops working again. :/ Should I send around scontrol 
> outputs? And yes, I watched out to set the --mem flag for the job 
> submission this time.

Well there you've said that the GPUs are only accessible by those 
numbered cores, and the first non-GRES job on the node will likely have 
taken them already.

You probably want to include all the cores on the appropriate sockets 
for the GPUs.

All the best,
Chris
-- 
   Chris Samuel  :  http://www.csamuel.org/  :  Berkeley, CA, USA