Talk:Running Serial Jobs on Niagara

From SciNet Users Documentation
Jump to: navigation, search


Hello, thanks for the work you've done preparing these help files as they are quite useful. However, I've stumbled across what appears to be an inconsistent passage:

  • The parameter -j $SLURM_TASKS_PER_NODE tells parallel to run 40 subjobs simultaneously on each of the nodes (note: do not use the similarly named variable $SLURM_TASKS_PER_NODE as its format is incompatible with GNU parallel).

The parallel manual specifies that the -j option controls the number of jobslots to use on each machine. Thus, I think the correct option is -j $SLURM_NTASKS_PER_NODE and not -j $NCORES (used in the code example) or -j $SLURM_TASKS_PER_NODE (used in the passage). The parenthetical note seems to be correct but is currently difficult to comprehend because of this typo.