On November XX Atos AD complex was updated with Slurm 22.05 and the same version of Slurm was installed on AC complex on .
AD has been default cluster (hpc-login and hpc-batch are nodes on AD) since .
The change to the new version impacts all Slurm jobs setting the number of OMP threads using directive:
#SBATCH --cpus-per-task
According to the official release note:
Srun will no longer read in SLURM_CPUS_PER_TASK. This means you will implicitly have to specify --cpus-per-task on your srun calls, or set the new SRUN_CPUS_PER_TASK env var to accomplish the same thing.
Consequently, users should adjust the jobs script by exporting SRUN_CPUS_PER_TASK environmental variable manually:
export SRUN_CPUS_PER_TASK=${SLURM_CPUS_PER_TASK:-1}
or by specifying number of OMP threads as an option with run command:
srun --cpus-per-task
To reduce user impact HPC support team has set:
export SRUN_CPUS_PER_TASK=${SLURM_CPUS_PER_TASK:-1}
to the user profile. However, user profile is automatically loaded only in batch jobs (with first line):
#!/bin/bash
In ksh, sh, and any other job type, one need to call user profile manually to benefit from the patch created by the HPC support team.
. /etc/profile
${SLURM_CPUS_PER_TASK: