[GE users] qrsh problem

reuti reuti at staff.uni-marburg.de
Mon Aug 30 21:10:09 BST 2010


Am 30.08.2010 um 20:11 schrieb deadline:

>>> <snip>
>>> I submit the job:
>>> 
>>> qrsh  -l h=norbert  hostname
>>> 
>>> qstat shows:
>>> 
>>> 584 0.50000 hostname   deadline     r     08/30/2010 12:45:53
>>> cluster at norbert
>>> 
>>> Then I qdel it:
>>> 
>>> qmaster messages:
>>> 
>>> 08/30/2010 12:46:55|worker|norbert|W|job 584.1 failed on host norbert
>>> assumedly after job because: job 584.1 died through signal HUP (1)
>>> 
>>> execd messages on norbert:
>>> 
>>>> 08/30/2010 12:46:55|  main|norbert|W|reaping job "584" ptf complains:
>>> Job does not exist
>>> 08/30/2010 12:46:55|  main|norbert|E|can't open file
>>> active_jobs/584.1/error: No such file or directory
>> 
>> As other jobs are working, that spool directory doesn't seem to be full or
>> write protected.
> 
> No, plenty of space, the job writes other data like the trace file
> local jobs work fine and write to the directory.
>> 
>> Can you qrsh "local" - from the headnode to itself?
> 
> yes, works fine.

Maybe it was in the discussion before: an ssh (outside of SGE) from any node to the headnode is working too?

-- Reuti

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=278300

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list