[slurm-users] Job failure issue in Slurm

[navin srivastava]

Hi Team,

i am seeing a weird issue in my environment.
one of the gaussian job is failing with the slurm within a minute after it
go for the execution without writing anything and unable to figure out the
reason.
The same job works fine without slurm on the same node.

slurmctld.log

[2020-06-03T19:14:33.170] debug:  Job 1357498 has more than one partition
(normal)(21052)
[2020-06-03T19:14:33.170] debug:  Job 1357498 has more than one partition
(normalGPUsmall)(21052)
[2020-06-03T19:14:33.170] debug:  Job 1357498 has more than one partition
(normalGPUbig)(21052)
[2020-06-03T19:15:12.497] debug:  sched: JobId=1357498. State=PENDING.
Reason=Priority, Priority=21052.
Partition=normal,normalGPUsmall,normalGPUbig.
[2020-06-03T19:15:12.497] debug:  sched: JobId=1357498. State=PENDING.
Reason=Priority, Priority=21052.
Partition=normal,normalGPUsmall,normalGPUbig.
[2020-06-03T19:15:12.497] debug:  sched: JobId=1357498. State=PENDING.
Reason=Priority, Priority=21052.
Partition=normal,normalGPUsmall,normalGPUbig.
[2020-06-03T19:16:12.626] debug:  sched: JobId=1357498. State=PENDING.
Reason=Priority, Priority=21052.
Partition=normal,normalGPUsmall,normalGPUbig.
[2020-06-03T19:17:12.753] debug:  sched: JobId=1357498. State=PENDING.
Reason=Priority, Priority=21052.
Partition=normal,normalGPUsmall,normalGPUbig.
[2020-06-03T19:18:12.882] debug:  sched: JobId=1357498. State=PENDING.
Reason=Priority, Priority=21052.
Partition=normal,normalGPUsmall,normalGPUbig.
[2020-06-03T19:19:13.633] sched: Allocate JobID=1357498 NodeList=oled4
#CPUs=4 Partition=normal
[2020-06-04T12:25:36.961] _job_complete: JobID=1357498 State=0x1 NodeCnt=1
WEXITSTATUS 2
[2020-06-04T12:25:36.961]  SLURM Job_id=1357498 Name=job1 Ended, Run time
17:06:23, FAILED, ExitCode 2
[2020-06-04T12:25:36.962] _job_complete: JobID=1357498 State=0x8005
NodeCnt=1 done

slurmd.log

[2020-06-04T12:22:43.712] [1357498.batch] debug:  jag_common_poll_data:
Task average frequency = 2769 pid 64084 mem size 4625724 23696420 time
164642.840000(164537+105)
[2020-06-04T12:23:13.712] [1357498.batch] debug:  jag_common_poll_data:
Task average frequency = 2769 pid 64084 mem size 4625724 23696420 time
164762.820000(164657+105)
[2020-06-04T12:23:43.712] [1357498.batch] debug:  jag_common_poll_data:
Task average frequency = 2769 pid 64084 mem size 4625724 23696420 time
164882.810000(164777+105)
[2020-06-04T12:24:13.712] [1357498.batch] debug:  jag_common_poll_data:
Task average frequency = 2769 pid 64084 mem size 4625724 23696420 time
165002.790000(164897+105)
[2020-06-04T12:24:43.712] [1357498.batch] debug:  jag_common_poll_data:
Task average frequency = 2769 pid 64084 mem size 4625724 23696420 time
165122.770000(165016+105)
[2020-06-04T12:25:13.713] [1357498.batch] debug:  jag_common_poll_data:
Task average frequency = 2769 pid 64084 mem size 4625724 23696420 time
165242.750000(165136+105)
[2020-06-04T12:25:36.955] [1357498.batch] task 0 (64084) exited with exit
code 2.
[2020-06-04T12:25:36.955] [1357498.batch] debug:  task_p_post_term:
affinity 1357498.4294967294, task 0
[2020-06-04T12:25:36.960] [1357498.batch] debug:
 step_terminate_monitor_stop signaling condition
[2020-06-04T12:25:36.960] [1357498.batch] job 1357498 completed with
slurm_rc = 0, job_rc = 512
[2020-06-04T12:25:36.960] [1357498.batch] sending
REQUEST_COMPLETE_BATCH_SCRIPT, error:0 status 512
[2020-06-04T12:25:36.961] [1357498.batch] debug:  Message thread exited
[2020-06-04T12:25:36.962] [1357498.batch] done with job
[2020-06-04T12:25:36.962] debug:  task_p_slurmd_release_resources: affinity
jobid 1357498
[2020-06-04T12:25:36.962] debug:  credential for job 1357498 revoked
[2020-06-04T12:25:36.963] debug:  Waiting for job 1357498's prolog to
complete
[2020-06-04T12:25:36.963] debug:  Finished wait for job 1357498's prolog to
complete
[2020-06-04T12:25:36.963] debug:  [job 1357498] attempting to run epilog
[/etc/slurm/slurm.epilog.clean]
[2020-06-04T12:25:37.254] debug:  completed epilog for jobid 1357498
[2020-06-04T12:25:37.254] debug:  Job 1357498: sent epilog complete msg: rc
= 0

any suggestion will be welcome to troubleshoot this issue further.

Regards
Navin.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.schedmd.com/pipermail/slurm-users/attachments/20200604/f9a4814f/attachment.htm>