Jobs missing from queue


Issue:  Sometimes jobs that were in a Queued state can disappear from the system.  This happens infrequently, but is most likely to happen after an upgrade, or sometimes even after a restart of pbs_server. 

 

Solution:  The easiest way to handle missing jobs is to resubmit the jobs with the script information that is saved by Torque.  In /var/spool/torque/server_priv/jobs/ you should see files for the jobs on the system.  There will usually be .JB and .SC files for each of the jobs.  You can also see files of other types for job arrays.  The .SC file contains the job script and can be used to re-submit any missing jobs. 

 

You will want to submit the script as the same user, which you can do if you're the root user with the -P option of qsub. 

Example: 

[root@support-machine1 ~]# qsub -P rsmith /var/spool/torque/server_priv/jobs/18087.support-bb.SC
18089.support-bb

Tags: disappear, gone, job, missing
Last update:
2015-10-19 18:44
Author:
Ben Roberts
Revision:
1.0
Average rating:0 (0 Votes)

You can comment this FAQ

Chuck Norris has counted to infinity. Twice.

Records in this category

Tags