Sean Dilda agrajag at dragaera.net
Wed Jun 16 19:23:14 BST 2004

I have a cluster running SGE 5.3.p5 and am looking to upgrading it to
6.0.   One of the things we have is high and low priority queues.  We
implement these by using the priority settings in the queue.  Low
priority queues get a nice value of 19, and high priority queues get a
nice value of -1.  It works quite well.

I tried to setup something similar on my test cluster that has SGE 6.0. 
I submitted some jobs to the low priority queues and found that their
nice values were changing.  I did some digging around and found
references to automatic job reprioritization.  I also found that it
could be disabled by setting execd_params to 'NO_REPRIORITIZATION' in
the sge_conf.  I did this, and the problem didn't go away.  I then
restarted sge_execd, and for good measure, sge_qmaster.  It seems to
still be doing the reprioritization.

Am I missing something?  Or is this a bug in SGE 6.0?



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