Re: [Condor-users] shadow exited

[Erik Paulson <epaulson@xxxxxxxxxxx>]

On Tue, Jun 15, 2004 at 03:54:06PM -0500, Scott Koranda wrote:
> Hi,
> 
> I found this in the SchedLog:
> 
> 6/15 12:48:11 Shadow pid 25284 for job 584394.0 exited with status 107
> 6/15 12:48:11 Called send_vacate( <129.89.200.23:41351>, 443 )
> 6/15 12:48:11 Sent RELEASE_CLAIM to startd on <129.89.200.23:41351>
> 6/15 12:48:11 Match record (<129.89.200.23:41351>, 584394, 0) deleted
> 6/15 12:48:11 Capability of deleted match: <129.89.200.23:41351>#1298467496
> 6/15 12:48:11 Entered delete_shadow_rec( 25284 ) 6/15 12:48:11 Deleting shadow rec for PID 25284, job (584394.0)
> 6/15 12:48:11 Entered check_zombie( 25284, 0x92af31c, st=2 )
> 6/15 12:48:11 Marked job 584394.0 as IDLE
> 
> Why might have the shadow for job 584394.0 exited with status
> 107?
> 

We want the job to go back into the queue. It's normal.  

Shadow exit codes should be considered a black box - we don't
follow UNIX conventions of exiting with '0' for success and
'1' for failure - we've got about 10 or so exit codes, all of
which are for success.

Is there something else going that you're worried about?

-Erik