The queue scheduler of ZEUS is SLURM (analogue of PBS/torque scheduler we use on Pluto and Mercury):
- Slurm documentation is available here.
- Basic sample submission scripts are listed HERE.
- Submission of the job via slurm script “scriptname.sh” can done with command “sbatch”:
sbatch scriptname.sh
- Queue states:
- Listing of running jobs:
- Delete job 1234 from queue
scancel 1234
- You can run any executable, e.g. “myexecutable”, on the zeuse’s nodes, e.g. zeus66 and zeus01…zeus06 without any slurm submission scripts (job will be placed in queue anyway):
srun -w zeus[66,101-106] ./myexecutable
Note: You will NOT be able to login to the compute node(s) from the head node if you haven’t jobs running on that compute node(s).
0 Comments.