[GE users] How to configure SGE

Reuti reuti at staff.uni-marburg.de
Thu Aug 3 12:47:08 BST 2006


Am 01.08.2006 um 10:35 schrieb Yoshio Tanaka:

> Hi,
> Would someone give advices and/or recommendations on how to configure
> SGE to satisfy the following requirements?
> - Each user is able to execute at least 24 simultaneous jobs if node
>   is available.
> - Even if a user is executing 24 jobs, he/she is allowed to execute
>   more jobs if node is available.

why just 24? Do you have enough nodes, that you could give each user  
24 machines/slots as his primary machines/slots, and later on use a  
type of secondary queue for each one?

> - If a user (user A) is executing more than 24 jobs and the other user
>   (user B) submit a new job, user A's excessive jobs will be killed
>   and user B's jobs will be activated.  User A's killed jobs will be
>   re-submitted to the queue.

To achieve this, you could combine a subordinate queue (to suspend  
the jobs) with the checkpointing feature, where a suspend will  
reschedule a job. But it's not a good setup, if nodes are hard-wired  
to users already as mentioned before.

A simple fair-share setup isn't working for you - are the jobs  
running a long time?

-- Reuti

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