Example job submission scripts for HPC systems

Contains PBS and SLURM examples

#####PBS

• To submit a single job from a *.sh file: qsub example-pbs-script.sh
• To submit multiple jobs in a loop: ./submit-pbs-loop.sh or python submit-pbs-loop.py (These do the same thing, but the Python example is easier to work with).

Note that the latter option does not require a *.sh file to exist for each job. It creates job submission strings and submits them without saving to a file. This is useful for large submissions.

pbs-serial-batch/ contains a few options for running lots of similar jobs. If your cluster supports array scheduling, this may be easier:

qsub serial-job.sh -t 0-100%10

This will submit 100 serial jobs, running a maximum of 10 at a time (the slot_limit parameter after the %). The array ID is passed into serial-job.sh via the $PBS_ARRAYID environment variable, which can be used to change what work is being done.

#####SLURM The SLURM examples are connected to each other, because the "multiple-submission" case requires a *.sh file to exist for the job (as far as I know).

• To submit multiple jobs in a loop: ./submit-slurm-loop.sh
• To submit a single job, this example would require you to pass an environment variable like so:

sbatch -J my_job_name --export=VARIABLE=1 example-slurm-script.sh

The job name is optional. If you modify example-slurm-script.sh to not depend on ${VARIABLE}, then you could just run sbatch example-slurm-script.sh.