Re: [Condor-users] Behavior of Condor jobs held for file transfer errors

[Todd Tannenbaum <tannenba@xxxxxxxxxxx>]

On 6/20/2012 3:00 PM, Steven C Timm wrote:
> What about held grid universe jobs which first require a
> Condor_rm
> And then a
> Condor_rm -forcex
> Any way to do that with a SYSTEM_PERIODIC_REMOVE?
>
> Steve Timm

Do you always want to simply remove held grid jobs?

If so, you can put the following into the submit file of a grid universe 
job:

   +nonessential = true

This tells Condor to simply abort (remove) any problematic job instead 
of putting the job on hold.  Condor will try to remove it nicely, but 
will not let it stick around in the queue even if it fails to confirm 
what happened on the execute node.  So placing the nonessential 
attribute in the job ad is equal to doing condor_rm followed by 
condor_rm -forcex anytime the job would have otherwise gone on hold.

regards,
Todd



> -----Original Message-----
> From: condor-users-bounces@xxxxxxxxxxx [mailto:condor-users-bounces@xxxxxxxxxxx] On Behalf Of Dan Bradley
> Sent: Wednesday, June 20, 2012 2:56 PM
> To: condor-users@xxxxxxxxxxx
> Subject: Re: [Condor-users] Behavior of Condor jobs held for file transfer errors
>
>
> Removing jobs that are on hold can be achieved by using the periodic_remove expression in the job submit file or by the SYSTEM_PERIODIC_REMOVE expression in the submit machine condor configuration.
>
> Example:
>
> SYSTEM_PERIODIC_REMOVE = HoldReasonCode == 12 || HoldReasonCode == 14
>
> The HoldReasonCodes are defined in the manual:
>
> http://research.cs.wisc.edu/condor/manual/v7.6/10_Appendix_A.html#82773
>
> --Dan
>
> On 6/20/12 12:25 PM, Myung Cho wrote:
> > Hi , I did a quick search for this topic but haven't found any
> > relevant posts. Is there a way to change/specify the default behavior
> > in Condor for jobs with file transfer errors? Our jobs with any error
> > in file transfer, for example a missing file specified in
> > transfer_output_files, seem to cause the job to be in held state for
> > ever. Is there a way for the job to just complete with error? I rather
> > see it finish with an error reported rather than have it just hang
> > around in hold state.
> >
> > Thanks.
> > _______________________________________________
> > Condor-users mailing list
> > To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx
> > with a
> > subject: Unsubscribe
> > You can also unsubscribe by visiting
> > https://lists.cs.wisc.edu/mailman/listinfo/condor-users
> >
> > The archives can be found at:
> > https://lists.cs.wisc.edu/archive/condor-users/
>
>
> _______________________________________________
> Condor-users mailing list
> To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
> subject: Unsubscribe
> You can also unsubscribe by visiting
> https://lists.cs.wisc.edu/mailman/listinfo/condor-users
>
> The archives can be found at:
> https://lists.cs.wisc.edu/archive/condor-users/
> _______________________________________________
> Condor-users mailing list
> To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
> subject: Unsubscribe
> You can also unsubscribe by visiting
> https://lists.cs.wisc.edu/mailman/listinfo/condor-users
>
> The archives can be found at:
> https://lists.cs.wisc.edu/archive/condor-users/


--
Todd Tannenbaum <tannenba@xxxxxxxxxxx> University of Wisconsin-Madison
Center for High Throughput Computing   Department of Computer Sciences
Condor Project Technical Lead          1210 W. Dayton St. Rm #4257
Phone: (608) 263-7132                  Madison, WI 53706-1685