de.jpgDE

SQUEUE

[user@login ~]$ squeue
JOBID PARTITION     NAME     USER ST       TIME  NODES NODELIST(REASON)
14726       big     LSUB    user1 PD       0:00    140 (Resources)
14743       big   job-01    user2 PD       0:00     90 (Priority)
14821       big      BTP    user3 PD       0:00     40 (Dependency)
14824       big      act    user3 PD       0:00     30 (Resources)
14827       big    sqt10    user4 PD       0:00     70 (Resources)
14710       big   RGB_D4    user4  R 4-02:36:20      1 c116
14875 sw01_shor  testRun    user   R       0:05      1 c002
14411 sw04_long    MISLC    user6 PD       0:00      4 (Dependency)
13212 sw04_long    MISLC    user6  R 9-21:22:05      4 c[088,107-109]
14437 sw04_long      z_t    user5  R    3:48:26      7 c[045-046,065...

Information shown in the partition display

The command squeue or its variants (e. g. squeue_all) provides information about all waiting and running jobs in the Workload Manager. This includes information such as time limit, requested number of nodes, queue assignment, etc…. A short explanation of the displayed columns is given in the following table.

ColumnExplanation
JOBIDis a unique identifier of a job after it has been successfully submitted with sbatch or similar.
PartitionDisplays the queue to be used.
NAME & USERDisplays the owner and description of a job.
ST State. Displays the status of a job. R / for Running, the job is running. PD for pending, means the job is waiting. CG for completing, the job is finished and only cleanup operations are executed
Nodelist (REASON)Displays the allocated nodes or displays states of waiting jobs.
BeginTime. There is a date before which the job may not start.
Dependency. The job waits for one or more jobs to finish before it starts itself.
JobHeld. The job is in hold status and must not start.
ReqNodeNotAvail. The requested resources are not available in the requested partition. e. g. the requested job time exceeds the time limit of the partition.
Resources /. The system waits for the requested resources.
TIMEShows the requested „Walltime“
TIME_LEFTHow much time is left in the running job before it is killed
NODESHow many nodes are required at least. Attention: An interval can also be requested, e. g. 5-10 nodes. Then only 5 nodes are shown here.
MIN_CPUHow many cores per node are requested.
NIC(E)Displays the nice value of a job. It is set by sbatch -nice 3. Nice lowers the priority of an own job.
MEMDisplays the required amount of RAM. If possible, always specify a limit ( #SBATCH -mem 120000MB), see HPC etiquette
START_TIMEPlanned start time of a job. Only the start times for the first job in the queue will be calculated.
A useful alias derived from the squeue_all command. This must be added to the. bashrc file in the home directory.
alias sq='squeue -o "%6P %6i %8u %.10j %.5D %.4c %11L %.11l  %3y %2h %2t %20S %20R" -S P,t,p'
[user@login ~]$ sq

PARTIT JOBID  USER NAME NODES MIN_ TIME_LEFT     TIMELIMIT  NIC SH ST START_TIME           NODELIST(REASON)
big    11495  ogoe jjpj    10   16 4-20:00:00   4-20:00:00  0   no PD 2016-08-07T23:05:52  (Resources)
big    11494  ogoe jjpj    10   16 4-20:00:00   4-20:00:00  0   no PD 2016-08-07T23:05:52  (Resources)
big    11493  ogoe jjpj    10   16 4-20:00:00   4-20:00:00  0   no PD 2016-08-07T23:05:52  (Resources)
big    11492  ogoe jjpj    10   16 4-20:00:00   4-20:00:00  0   no PD 2016-08-07T23:05:52  (Resources)
big    11408  zieg mamm    30   16 1-00:00:00   1-00:00:00  0   no PD 2016-08-05T10:11:53  (Resources)
big    11228  seen BTPu    40   16 4-19:00:00   4-19:00:00  0   no PD N/A                  (Dependency)
big    11171  jdor job-    90   16 3-00:00:00   3-00:00:00  0   no PD 2016-08-07T23:05:52  (Priority)
big    11056  rozi 13LS   140   16 4-10:00:00   4-10:00:00  0   no PD 2016-08-08T23:49:18  (Resources)
big    11490  misr mixt     4   16 6:47:41        12:00:00  0   no R  2016-08-04T15:48:31  c[158-161]
big    11395  mart RGB_     1   16 11:52:19     4-20:00:00  0   no R  2016-07-31T12:53:09  c116
big    11245  jvog phon    15   16 4-02:48:28   4-20:00:00  0   no R  2016-08-04T03:49:18  c[040-044,144-150,15
big    11227  seen BTPu    40   16 3-02:05:02   4-19:00:00  0   no R  2016-08-03T04:05:52  c[005-014,117-128,13
big    11110  ogoe jjpn    70   16 13:11:03     4-20:00:00  0   no R  2016-07-31T14:11:53  c[015-022,027,029-03
sw04_l 11485  nlic MISK     4   16 14-20:00:00 14-20:00:00  0   no PD N/A                  (Dependency)
sw04_l 11471  eshg albo     7   16 13-08:40:13 13-18:00:00  0   no R  2016-08-04T11:41:03  c[045-046,065-066,08
sw04_l 11242  nlic MISK     4   16 4-17:06:34  14-20:00:00  0   no R  2016-07-25T18:07:24  c[088,107-109]


guide/squeue.txt · Zuletzt geändert: 2018/05/14 20:59 von Sebastian Engel
Nach oben
CC Attribution-Share Alike 3.0 Unported
Driven by DokuWiki Recent changes RSS feed Valid CSS Valid XHTML 1.0