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

Re: [Condor-users] rebooting a submission node



On Tue, 17 May 2005 14:07:14 -0500  Erik Paulson wrote:

> Jobs that have been submitted with a JobLeaseDuration set will
> survive a reboot of the submit machine.

that's only true if you do a "hard" shutdown of the schedd, first.
namely:

condor_off -fast -schedd

if you do a graceful shutdown (SIGTERM, condor_off, etc), condor tries
to evict jobs, and in the case of vanilla jobs, that means killing
them and restarting them.

> Anything jobs without the job lease duration will be checkpointed
> (if possible) and then resumed or restarted when the condor_schedd
> starts running again, but will not be running while the submit
> machine is down.

right.

-derek