[slurm-users] ntasks-per-node for mpirun

[Mahmood Naderan]

Hi
With "cpu=48,mem=40G" limits and the following script

#!/bin/bash
#SBATCH --job-name=gr
#SBATCH --output=my_gr.log
#SBATCH --partition=SEA
#SBATCH --account=fish
#SBATCH --mem=6GB
#SBATCH --nodes=4
#SBATCH --ntasks-per-node=20
mpirun -np $SLURM_NTASKS /share/apps/gromacs-2019.6/single/bin/gmx_mpi
mdrun -v -ntomp 1 -deffnm nvt


I see that job goes to the pending state with the AssocGrpCpuLimit reason.
I expected to have 4 nodes each running 5 mpi processes.
Isn't that? So why has the CPU limit been exceeded?


Regards,
Mahmood
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.schedmd.com/pipermail/slurm-users/attachments/20200928/3e73be01/attachment.htm>