On a compute node with multiple slots, are the running jobs each explicitly assigned a slot ID as they start, and if so how can the user or submission script see it?
To see the job ID, one can use the $JOB_ID environment variable within the submission script. What about the slot number?
I looked for slot information using qstat -j <job number> but the information about the job does not contain any information about which of the slots the job is using. I was hoping there would be an integer variable related to the slot number.
EDIT: in the general case, a job might be assigned multiple slots if it is parallelized, so in this case the list of slot IDs could be determined.
