To use these,
- clone the repository at, say,
~/.software/cluster-utils/; and - depending on the scheduler available, add one of the following to your
~/.bash_profile.
_job_script_name="job.pbs"
_cluster_utils_location="$HOME/.software/cluster-utils"
. "$_cluster_utils_location"/bash/pbs/monitoring.sh
. "$_cluster_utils_location"/bash/pbs/submission.sh
. "$_cluster_utils_location"/bash/submission_common.sh
_job_script_name="job.sh"
_cluster_utils_location="$HOME/.software/cluster-utils"
. "$_cluster_utils_location"/bash/slurm/monitoring.sh
. "$_cluster_utils_location"/bash/slurm/submission.sh
. "$_cluster_utils_location"/bash/submission_common.sh
ja: show all jobsjaq: show all queued jobsjar: show all running jobsmyq: show current user's jobsmyj: more detailed display of current user's jobs
jdel $job_id: cancels a jobjint: requests an interactive jobjhold $job_id: hold a queued jobjrls $job_id: release a held jobsub $job_dir_1 $job_dir_2: submit jobs, one job for each given directorysubh $job_dir_1 $job_dir_2: submit jobs in a held state
pc_stop $job_id_1 $job_id_2pc_reload $job_id_1 $job_id_2pc_sub $job_id_1 $job_id_2pc_subh $job_id_1 $job_id_2