Seitenvergleich

...

1. Ask for an interactive shell with the command srun <…> --pty --interactive bash. We For reduced waiting times we advise to use the one of the test partitions for when submitting interactive jobs.
2. In the interactive shell, execute the parallel program with the MPI starter mpirun or srun.

blogin1:~ > srun -t 00:10:00 -p medium40:test -N2 --tasks-per-node 24 --pty --interactive bash
bash-4.2$ mpirun hello_world >> hello_world.out
bash-4.2$ exit
blogin1:~ > 

...

#!/bin/bash
#SBATCH -t 00:20:00
#SBATCH -N 1
#SBATCH --cpus-per-task=96
#SBATCH -p large96:test

# This binds each thread to one core
export OMP_PROC_BIND=TRUE
# Number of threads as given by -c / --cpus-per-task
export OMP_NUM_THREADS=$(($SLURM_CPUS_PER_TASK * 2))
export KMP_AFFINITY=verbose,scatter

hello_world > hello.output

Job Accounting

The webpage Accouting Accounting and NPL gives you more information about job accounting.

...