Submitting Batch Jobs
The table below displays the types of commands needed for batch processing on most Research Computers at VCU.
Locate the type of command and the machine you are using.
| Activity | Machine | Command | Example |
| Sends a job to the batch queues allowing you to log off. There are three batch queues available, "short", "normal" and "night". The default is normal. Always use this command to run any jobs over a few minutes long. This allows everyone to equally share the computer resources. | Edison(Xena) | bsub | bsub sas mysasjob.sas bsub -qshort sas mysasjob.sas bsub -qnight sas mysasjob.sas |
| Aurora | qsub | qsub
sas mysasjob.sas qsub -qshort sas mysasjob.sas qsub -qnight sas mysasjob.sas |
|
| Lists the batch queues. Please note Hydras batch queues: 4 processor, limit 1 job/user named p4p 2 processor, limit 2 jobs/user named p2p 1 processor, limit 4 jobs/user named serial |
Edison(Xena) | bqueues | |
| Aurora | qstat | qstat
-q Lists the available queues and their limits. qstat -s Lists the jobs and any comments provided by the batch system or administrator. |
|
| Checks the status of a batch job. Provides information about jobs that have been submitted. | Edison(Xena) | bjobs | bjobs -- shows basic information |
| Aurora | qstat | qstat -- shows only basic information qstat -f -- shows full information qstat -f -Q -- shows full information about all the available queues qstat -f -Q test -- shows full information about the queue 'test' |
|
| Kills a batch job submitted with bsub | Edison(Xena) | bkill | |
| Aurora | qdel | qdel 1234 -- where '1234' is the job ID, which can be obtained by the command qstat. | |
| Stops a job submitted with bsub | Edison(Xena) | bstop | |
| Aurora | A job will be stopped when its execution time reaches either the time limits set in the submission-script or the time limits of the queue on which the job runs. | ||
| Resumes a previously stop job from "bstop" | Aurora, Xena |
bresume | |
| If your job will not finish before PBS stops it, then you should write your program so that it can checkpoint itself when it is stopped. You have to make two changes: you have to modify your program and your PBS submission script. | |||
