Re: [Condor-users] condor_rm on lots of jobs

[Ian Stokes-Rees <ijstokes@xxxxxxxxxxxxxxxxxxx>]

On 4/29/10 12:31 PM, Greg Thain wrote:
> If you aren't using DAGman, one quick and dirty way to work around
> this problem, is to remove all jobs in two steps:  first remove all
> idle jobs:
>
> condor_rm -const 'JobStatus == 1'

We are using DAGMan.  Is there anything special we should do for these
condor_rm's, other than just remove the DAG jobs, and let DAGMan take
care of the rest?

Also, for the two-stage remove you propose, how long should we wait in
between? Until 50%, 90%, 99%, 100% of the idle jobs are removed?

I've summarized the input I've received from people on this page:

http://portal.nebiogrid.org/docs/condor/condor_rm

I'm still hoping for some guidance on how to set up multiple schedd's.

Thanks,

Ian