...
Code Block |
---|
bash-4.2$ sbatch test_slurm.sh Submitted batch job 8501 |
To retrieve the information about the submitted jobs you can use the command "squeue":
Code Block |
---|
bash-4.2$ squeue
JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON)
8501 slurmHPC_ test_sl apascoli R 0:01 1 hpc-200-06-05 |
The table shows 8 columns: JOBID, PARTITION, NAME, USER, ST, TIME, NODES and NODELIST(REASON).
- JOBID shows the id corresponding to the submitted jobs.
- PARTITION which in slurm is a synonym of "queue" indicates to which partition the node belongs.
- NAME it corresponds to the name assigned in the submit file, otherwise it will match the name of the submit file.
- USER indicates the user who submitted the job.
- ST indicates if the jobs are running ("R") or if it is pending ("PD").
- TIME shows how long the jobs have run for using the format days-hours:minutes:seconds.
- NODES indicates the number of machines running the job.
- NODELIST indicates where the job is running or the reason it is still pending.
Then, to see information about a single job use:
Code Block |
---|
bash-4.2$ sstat --format=JobID,AveCPU -j 8501
JobID AveCPU
------------ ----------
8501.0 213503982+ |
The option --format allows to customise the output based on the desired features.
For instance in the example above are shown:
- JobID
- AveCPU Average (system + user) CPU time of all tasks in job.
Many more features are listed in the slum manual. [31]
The output, from the option --output=tasks1.txt, should be something like (hostnames and formatting may change):
...