[GE users] Jobs disappering from SGE: shepherd exited with exit status 19

Ondrej Bojar bojar at ufal.ms.mff.cuni.cz
Mon Nov 26 12:07:45 GMT 2007


    [ The following text is in the "ISO-8859-1" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]

Dear Goncalo,

I'm not experienced enough yet, but I was getting similar errors (shepherd 
exited with exit status 19) when a firewall got turned on during some automatic 
updates. Our admin first insisted that he has not changed anything in the setup 
but he later discovered this was the reason.

I'm curious what more informed users or authors will tell us.

Best luck, Ondrej.

Gon?alo Borges wrote:
> 
> Dear All,
> 
> We are using sge60u7_1 and we want to report the following reason.
> 
> - Because of administrative issue, we had to suspend some user jobs last 
> week.
> - The jobs were properly suspended and still appeared on the queue last 
> Thursday, but today, when we returned to the office, the jobs just 
> disappeared from the system.
> - For one of these jobs, we see the following message in Qmaster log:
> 
> 11/22/2007 16:23:06|qmaster|sge01|W|job 132286.1 failed on host 
> lfcomp02.lip.pt before writing exit_status because: shepherd exited with 
> exit status 19
> 
> - On the Exec node, we see the following Log:
> 
> 11/22/2007 16:23:04|execd|lfcomp02|I|controlled shutdown 6.0u7
> 11/22/2007 16:23:05|execd|lfcomp02|I|starting up 6.0u7
> 11/22/2007 16:23:05|execd|lfcomp02|E|abnormal termination of shepherd 
> for job 132286.1: "exit_status" file is empty
> 11/22/2007 16:23:05|execd|lfcomp02|E|can't open usage file 
> "active_jobs/132286.1/usage" for job 132286.1: No such file or directory
> 11/22/2007 16:23:05|execd|lfcomp02|E|shepherd exited with exit status 19
> 11/22/2007 16:23:06|execd|lfcomp02|I|PTF_MAX_PRIORITY=0, 
> PTF_MIN_PRIORITY=20
> 
> Can anyone shed some light on this issue?
> 
> Thanks in Advance
> Cheers
> Goncalo
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
> For additional commands, e-mail: users-help at gridengine.sunsource.net

-- 
Ondrej Bojar (mailto:obo at cuni.cz / bojar at ufal.mff.cuni.cz)
http://www.cuni.cz/~obo

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list