Job submission
The general use of SLURM on the cluster is described here (RWTH High Performance Computing Linux).
The only difference between the submission to a public partition and the submission to the maintenance partition is the specification of the project. For example, the following script is submitted for the standard partition (c18m):
normal.sh |
#!/usr/bin/zsh #SBATCH --job-name=my_job #SBATCH --output=output_%J.txt #SBATCH --mem-per-cpu=512M #SBATCH --time=60 a.out |
By specifying a suitable project (accounts), the job is submitted to the IH partition.
| ih.sh |
#!/usr/bin/zsh #SBATCH --job-name=my_job #SBATCH --output=output_%J.txt #SBATCH --mem-per-cpu=512M #SBATCH --time=60 #SBATCH --account=abcd a.out |
