[GE users] Scheduling of long jobs

Vladimir vlists at veus.hr
Fri Aug 20 00:27:36 BST 2004


Patrice Seyed wrote:

>Interesting Reuti, but like you said a job in long00b stay there even if a
>long00b is open. I'm not sure if this method is "smarter" than mine, but
>with your method you don't have a possible scenerio where 3 jobs are running
>over 2 cpus, but even though that can occur on mine, it can for no more than
>2 hours (hard limit on express queues), unless there is are jobs waiting to
>get into an express queue. 
>
>I agree it would be nice to be able to suspend a slot instead of a queue.
>The current setup is more attuned for single cpu jobs, and also for my
>cluster making a queue for each single cpu doesn't seem feasible.
>  
>

I will add a little complication in the whole setup involving FlexLM. 
Suspending a job outright may not release its license. It depends from a 
software to software and you should check whether the software you are 
using will release (check-in licenses) back to the FlexLM when 
suspended. Also we have noticed that some software that uses FlexLM has 
really hard time recovering when job is unsuspended ie. it is not able 
to reacquire a license.

You may be best off designating couple machines for "short" jobs and not 
allow any long jobs on it.

Vladimir

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list