[GE users] rerun queue?

gutnik gutnik at gmail.com
Fri Aug 27 22:03:51 BST 2010

Is it possible to set up SGE 6.2u4 to kill and resubmit some jobs when
(some load measure) exceeds a limit?

I could do this on a host-by-host basis (I think) by setting a suspend
threshold and setting the suspend method to be
something like kill -KILL. (Maybe I would need to do something with an
epilog script?) But I would like to do this
cluster-wide: if the number of free slots (or some other consumable)
drops below a certain point, I want to start killing
jobs in the rerunnable.q until I get back to a fixed point. Is there a way?



To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].

More information about the gridengine-users mailing list