[HTCondor-users] Force dag to continue after periodic_remove

["ASHIKUZZAMAN IDRISY" <aui109@xxxxxxx>]

Hi, 

Thank you very much, this solution worked. I should have thought of this, having re-checked the manual for DAGs (http://research.cs.wisc.edu/htcondor/manual/v8.0/2_10DAGMan_Applications.html), I found this line: "If the job fails and there is a POST script, node failure is determined by the exit value of the POST script. " 

Thus, I basically made a POST script that simply echo-ed something, and the subsequent jobs, B1-B600 were started. 

Thanks again,

Ash

Is there a way to force periodic_remove or something similar, to act like a
success in the eyes of the dag, so that subsequent jobs get completed?

You can add a POST script to your 20 A1... A20 node that turns a periodic remove into a success.  You can do something like this in your DAG file:

  Script Post A1 my_test.pl $JOBID

$JOBID will expand to the condor ID of the job.

So then your post script can do a condor_history command on the job -- if JobStatus is 3 that means removed -- I don't know if you can distinguish from a periodic remove vs. a "manual" remove...  You may need to do a little digging to figure that out.

Kent Wenger
CHTC Team

Ashikuzzaman Idrisy
Graduate Assistant
Pennsylvania State University