Re: [Condor-users] draining jobs from only one node

[Matt Hope <matthew.hope@xxxxxxxxx>]

On Thu, 10 Feb 2005 12:04:59 -0600, Gary Stiehr <stiehr@xxxxxxxx> wrote:
> 
> Erik Paulson wrote:
> 
> >On Wed, Feb 09, 2005 at 03:36:40PM -0600, Gary Stiehr wrote:
> >
> >
> >>Hello,
> >>
> >>We have a cluster running the following condor version:
> >>
> >># condor_version
> >>$CondorVersion: 6.6.6 Jul 26 2004 $
> >>$CondorPlatform: I386-LINUX_RH72 $
> >>
> >>
> >>We are interested in draining the jobs from just one node so that we can
> >>perform maintenance on it.  That is, we would like to allow the
> >>currently running jobs to finish but to not allow any further jobs to
> >>start on this node.  Is there a way to do that in this version of
> >>condor?
> >>
> >>
> >
> >Set START=FALSE. No new jobs will be admitted to the execute machine.
> >
> >
> >
> >
> Thanks, I will try that out.  I can see that it would clearly keep new
> jobs from being scheduled for this node.  However, would this cause
> existing jobs to be suspended or vacated once a condor_reconfig is done?

Not unless you reference the start expression itself in any other
statements like preempt

Matt