[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] Getting a job re-queued and exit script for cleanup after job exit



On Jul 21, 2005, at 3:33 PM, Terrence Martin wrote:

I was wondering if anyone could point me at the following.

The first is does condor provide a way to have an exit script executed after a job successfully completes? I am thinking a shell script that can do some more detailed post job cleanup would be useful.

Also is it possible for a job to send a message that indicates that it should be rescheduled to another job slot? The situation that could occur is that I have a user_wrapper that executes prior to the job and then runs the job via an exec call. Part of this wrappers job is to set up an environment specific to the owner of the job. If the environment cannot be setup for some reason however I would like to be able to signal to condor that it should try scheduling the job somewhere else. Is this possible?

You can use on_exit_remove to control whether a job leaves the queue or returns to idle state to be rerun. It can be an arbitrary boolean _expression_, but testing for a specific exit code is a common use:

== True || ExitCode != 13

With this _expression_, the job will leave the queue normally if it terminates with a signal or an exit code other than 13. If it terminates with exit code 13, it will be requeued and rerun.

+----------------------------------+---------------------------------+

|            Jaime Frey            |  Public Split on Whether        |

|        jfrey@xxxxxxxxxxx         |  Bush Is a Divider              |

|  http://www.cs.wisc.edu/~jfrey/  |         -- CNN Scrolling Banner |

+----------------------------------+---------------------------------+