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

Re: [Condor-users] gracefully remove nodes from pool



Erik Paulson wrote:

On Wed, Sep 22, 2004 at 03:29:33PM -0700, Jimmy M. Ferng wrote:


On Wed, 22 Sep 2004, Ian Chesal wrote:



Date: Wed, 22 Sep 2004 12:04:03 -0700
From: Ian Chesal <ICHESAL@xxxxxxxxxx>
Reply-To: Condor-Users Mail List <condor-users@xxxxxxxxxxx>
To: Condor-Users Mail List <condor-users@xxxxxxxxxxx>
Subject: RE: [Condor-users] gracefully remove nodes from pool

You want to use condor_off with the -peaceful flag which tells it to
wait indefinitly for running jobs to finish before turning off condor on
the machines. Like this:

condor_off -peaceful <machine1> [<machine2>...]



I was trying to do exactly the same thing as Andrew has asked about. I
tried:

condor_off -graceful -name <machine1> [<machine2>...]

It did remove schedd and startd on the remote machines. However, when
condor_startd was removed, the running job was also removed immediately.
(I am running Condor 6.6.5) I was hoping that I could keep the job
running until finished but not to take any more new jobs. Any advice?




condor_off -peaceful is new in 6.7.1 (note that -graceful and -peaceful do different things)




Yes the problem is when I tried this the job was killed. I am not running 6.7 so I do not have the -peaceful option