Ouessant : Control commands for batch jobs

Batch jobs are managed by the LSF software on all the nodes. The principal commands used to control your jobs are presented below.

To submit a job : bsub

The bsub command is used to submit a batch job:

$ bsub < mon_job.sh
Job <19661> is submitted to queue <computet1>

To display the waiting or executing jobs : bjobs

The bjobs command displays information about the evolution and consumption of all your batch jobs on the machine.

$ bjobs 
19661    Hydro          rlab432  RUN    Apr 19 11:09   53 second(s)     20     20*ouessantm03
19662    Hydro          rlab432  PEND         -        0 second(s)        -    -

The STAT (Status) column indicates if your job is running (RUN) or pending (PEND).

Attention : If you do not find the return of a finished job, it is very probable that you are not in the submission directory. It is also possible that you have exceeded the quota, or that you have not specified the output and error files in your submission script (lines #BSUB -o and #BSUB -e). In the latter case, your job return is lost.

To delete a job : bkill

The bkill command is used to delete a job. For example, to delete job 19661 which is running on one of the nodes of the machine:

$ bkill 19661
Job <19661> is being terminated

To see the outputs during execution: bpeek

The standard and error outputs are buffered during the job execution. Consequently, the files specified in the script remain empty during execution.

To display the outputs, it is necessary to use the bpeek <job_id> command. If <job_id> is not indicated, this command displays the outputs of the last executed job.