Skip to content

List Jobs by squeue & sview

squeue command

After you submit jobs, you can check their information and status in the queue. The simplest way is to use squeue command to list all of your jobs:

$ squeue -l -u $USER
Thu Aug  2 14:45:57 2018
             JOBID PARTITION     NAME     USER    STATE       TIME TIME_LIMI  NODES NODELIST(REASON)
              9405 general-s     Job1   nobody  RUNNING       0:32  01:00:00      1 lac-198
              9406 general-s Rmpi_tes   nobody  RUNNING       0:11     30:00      2 lac-[386-387]
              9290 general-l  LongJob   nobody  PENDING       0:00  36:00:00     40 (Resources)

where the argument -l reports more of the available information (i.e. l for long format) and -u specifies which user's jobs to show. You may find a complete squeue specifications from the SLURM web site.

HPCC staff also wrote some powertools commands so users can see their jobs conveniently. Before using powertools commands, please make sure powertools module is loaded by using module list (By default, the powertools module should be loaded unless the user has purged modules with module purge)

$ module list

Currently Loaded Modules:
  1) powertools/1.2

One of the commands sq works the same as the above squeue command:

$ sq                         # powertools command
Thu Aug  2 14:48:51 2018
             JOBID PARTITION     NAME     USER    STATE       TIME TIME_LIMI  NODES NODELIST(REASON)
              9405 general-s     Job1 username  RUNNING       0:35  01:00:00      1 lac-198
              9406 general-s Rmpi_tes username  RUNNING       0:14     30:00      2 lac-[386-387]
              9290 general-l  LongJob username  PENDING       0:00  36:00:00     40 (Resources)

where it shows the job IDs, job partition, job name, username, job state, elapsed time, walltime limit, total number of nodes and the node list or the waiting reason for the user's each jobs. Users can also use qs command to see more information:

$ qs                         # powertools command
Thu Aug  2 14:49:27 2019
                                                                                        Start_Time/
     JobID         User     Account     Name   Node CPUs  TotMem  Tres   WallTime  ST  Elapsed_Time  NodeList(Reason)
-----------------------------------------------------------------------------------------------------------------------
    60889405    MyHPCCAcc   general       Job1    1   16      2G  gpu:1    3:55:00  R      1:49:24   lac-198
    60889496    MyHPCCAcc   general  Rmpi_test    2    8   1500M   N/A       30:00  R        29:46   lac-[386-387]         
    60889290    MyHPCCAcc  classres    LongJob   40   80    750M  k80:1 3-00:00:00 PD 08-03T10:29:41 (Resources)

where more items, such as, total number of CPUs, memory, gpu per node and job start time (for pending jobs) or elapsed time (for running jobs) are shown. For a complete usage of squeue command, please refer to the SLURM web site.

sview command

Besides the text listing of the jobs, SLURM also offer a command to show the squeue information with a graphical interface. Use the command sview:

$ sview

You will see an image of a job list displaying all jobs in the queue:

A screenshot of the SLURM job queue graphic interface. On the left, a grid of red crosses show queued jobs, and on the right a tabbed window shows a list of job IDs and partitions.

Click on each job, it will pop out another window and show the detailed information. For a complete usage of sview command, please refer to the SLURM web site.