squeue

The ‘squeue’ command will show users the jobs that are currently scheduled:

[johnchris@l001 ~]$ squeue

JOBID     PARTITION   NAME     USER       ST   TIME  NODES NODELIST(REASON)
2197268   exclusive   jobOne   johnchirs  PD   0:00  12    (Priority)
2197266   exclusive   jobTwo   johnchris  PD   0:00  12    (Priority)
2197267   exclusive   jobThree johnchris  PD   0:00  12    (Priority)
2197265   exclusive   jobFour  johnchris  PD   0:00  12    (Resources)
2197269   exclusive   jobFive  johnchris  PD   0:00  12    (Priority)
2197270   exclusive   jobSix   johnchris  PD   0:00  12    (Priority)
2216941   shared      Job      johnchris   R   6:51:59     1  c132
2214572   gpua100     python1  johndoe     R   1-09:09:51  1  g012
2216584   gpuv100     python2  johndoe     R   1-09:06:16  1  g002
2216679   shared      python3  johndoe     R   22:35:27    1  c012
2216680   shared      matlab   johnchris   R   22:35:27    1  c014
2216677   shared      jupyter  johnchris   R   22:35:58    1  c002
JOBIDJob or step ID
PARTITIONName of a partition. shared, exclusive, gpu100 and gpuv100 etc.
NAMEName of the job in the queue
USEROwner of the job in the queue
STState of the job. Include: allocated, completing, down, drained, draining, fail, failing, future, idle
TIMETime used by the job. days-hours:minutes
NODESNumber of nodes allocated to the job
NODELIST (REASON)For pending and failed jobs, this field displays the reason for pending or failure. Otherwise, this field shows a list of allocated nodes

The squeue command gives us the following information:

  • JOBID: The unique ID for your job.
  • PARTITION: The partition your job is running on (or scheduled to run on).
  • NAME: The name of your job.
  • USER: The username for whomever submitted the job.
  • ST: The status of the job. The typical status codes you may see are:
    • CD (Completed): Job completed successfully.
    • CG (Completing): Job is finishing, Slurm is cleaning up.
    • PD (Pending): Job is scheduled, but the requested resources aren’t available yet.
    • R (Running): Job is actively running.
  • TIME: How long your job has been running.
  • NODES: How many nodes your job is using.
  • NODELIST(REASON): Which nodes your job is running on (or scheduled to run on). If your job is not running yet, you will also see one of the following reason codes:
    • Priority: When Slurm schedules a job, it takes into consideration how frequently you submit jobs. If you often submit many jobs, Slurm will assign you a lower priority than someone who has never submitted a job or submits jobs very infrequently. Don’t worry, your job will run eventually.
    • Resources: Slurm is waiting for the requested resources to be available before starting your job.
    • Dependency: If you are using dependent jobs, the parent job may show this reason if it’s waiting for a dependent job to complete.

“squeue -u username” will view only the jobs from a specific user [johnchris@l001 ~]$ squeue -u johnchris 

2213604   exclusive  JobI    johnchris  PD  0:00    2  (Priority)
2213606   exclusive  JobII   johnchris  PD  0:00    2  (Priority)
2213618   exclusive  JobIII  johnchris  PD  0:00    2  (Priority)
2213620   exclusive  JobIV   johnchris  PD  0:00    2  (Priority)
2213621   exclusive  JobV    johnchris  PD  0:00    2  (Priority)
2214144   exclusive  JobVI   johnchris  PD  0:00    2  (Priority)
2217039   shared     JobVII  johnchris  PD  0:00    5  (Priority)
2213574   shared     JobVIII johnchris  R   2-05:17:14  1  c021
2217608   shared     JobIX   johnchris  R   7:34:08     1  c106
2216796   shared     JobX    johnchris  R   13:19:33    7  c[116-122]

[johnchris@l001 ~]$ squeue -u johnchris --start

2197281   exclusive  A      johnchris  PD  N/A          6   (null)   (Priority)
2197283   exclusive  B      johnchris  PD  N/A          6   (null)   (Priority)
2197284   exclusive  C      johnchris  PD  N/A          6   (null)   (Priority)
2216608   exclusive  D      johnchris  PD  N/A         24   (null)   (Priority)
2216842   exclusive  E      johnchris  PD  N/A         36   (null)   (Priority)
2197265   exclusive  F      johnchris  PD  2024-04-12T07:33:39  12  c[163-174]  (Resources)
2197266   exclusive  G      johnchris  PD  2024-04-13T16:30:23  12  c[176-187]  (Priority)
2197267   exclusive  H      johnchris  PD  2024-04-13T16:30:23  12  c[188-199]  (Priority)
2197268   exclusive  I      johnchris  PD  2024-04-13T19:21:20  12  c[151-155,200-206]  (Priority)

Note: The job start time is an estimate; it may change as other jobs are submitted. The memory and wall clock time parameters require you to estimate the amount of time your job will need to finish. If you underestimate, your job may be killed before it completes. If you overestimate, your job may take longer to start, as the queuing system finds a way to allocate the resources.

Scroll to Top