[torqueusers] Per job epilogue script

Vincent LIARD vincent.liard at scilab.org
Tue Feb 16 02:25:44 MST 2010


Hello,

I don't manage to get a per job epilogue script to run. I've read
http://www.clusterresources.com/products/torque/docs/a.gprologueepilogue.shtml#perJobScripts

(Note that the example given in the documentation is contradictory since
"files must [have] at least execute and write privileges" but the
example shows 'r-x' files... Moreover the qsub call references
"epilogue=/home/usertom/dev/epilogue_script.sh" while the previously
listed files are said to be in "$PBS_HOME/mom_priv/".)

As for my problem :

$ ls -al epi.sh 
-rwx------ 1 ecurie ecurie 281 2010-02-16 09:56 epi.sh

$ cat epi.sh 
#!/bin/sh
echo "Job ID: $1"
touch /home/ecurie/epilogue.mark
exit 0

$ echo hostname | qsub -l epilogue=/home/ecurie/epi.sh
337.eccica

$ tracejob 337

Job: 337.eccica

02/16/2010 10:00:31  S    enqueuing into batch, state 1 hop 1
02/16/2010 10:00:31  S    Job Queued at request of ecurie at eccica, owner = ecurie at eccica, job name = STDIN, queue = batch
02/16/2010 10:00:31  S    Job Modified at request of Scheduler at eccica
02/16/2010 10:00:31  S    Exit_status=0 resources_used.cput=00:00:00 resources_used.mem=0kb resources_used.vmem=0kb resources_used.walltime=00:00:00
02/16/2010 10:00:31  L    Job Run
02/16/2010 10:00:31  S    Job Run at request of Scheduler at eccica
02/16/2010 10:00:31  M    job was terminated
02/16/2010 10:00:31  M    checking job post-processing routine
02/16/2010 10:00:31  S    dequeuing from batch, state COMPLETE
02/16/2010 10:00:31  M    obit sent to server
02/16/2010 10:00:31  A    queue=batch
02/16/2010 10:00:31  M    scan_for_terminated: job 337.eccica task 1 terminated, sid=8471
02/16/2010 10:00:31  A    user=ecurie group=ecurie jobname=STDIN queue=batch ctime=1266310831 qtime=1266310831 etime=1266310831 start=1266310831 owner=ecurie at eccica exec_host=ecurie0/0 Resource_List.epilogue=/home/ecurie/epi.sh
                          Resource_List.neednodes=1 Resource_List.nodect=1 Resource_List.nodes=1 Resource_List.walltime=01:00:00 
02/16/2010 10:00:31  A    user=ecurie group=ecurie jobname=STDIN queue=batch ctime=1266310831 qtime=1266310831 etime=1266310831 start=1266310831 owner=ecurie at eccica exec_host=ecurie0/0 Resource_List.epilogue=/home/ecurie/epi.sh
                          Resource_List.neednodes=1 Resource_List.nodect=1 Resource_List.nodes=1 Resource_List.walltime=01:00:00 session=8471 end=1266310831 Exit_status=0 resources_used.cput=00:00:00 resources_used.mem=0kb
                          resources_used.vmem=0kb resources_used.walltime=00:00:00

And I don't see any output from epilogue (either on stdout or in
STDIN.[eo]337) and I don't see an epilogue.mark file created.

Do you have any idea ?

-- 
Vincent LIARD
Ingénieur de développement
--
Consortium Scilab
Digiteo
Domaine de Voluceau
Rocquencourt - B.P. 105
78153 Le Chesnay Cédex
Tél.: 01.39.63.55.47




More information about the torqueusers mailing list