Dear Rike
I'm facing the same error in my own cluster, slurm version 23.11.3. And i notice that my task are running in sequence, not in parralele. I'm using the example inside the srun manual:
#!/bin/bash
srun -n1 sleep 30 & srun -n1 sleep 45 & srun -n1 sleep 20 & srun -n1 sleep 25 & wait
$ sbatch -n4 test.sh
I hope that should running in 45 seconds, as my server have 64 cores... But no, each task is running sequencialy.
Hope that someone could help us?
Regards