[GE users] Reservation and Backfilling with SGE 6.0 on a Linux cluster

Korambath, Prakashan ppk at ats.ucla.edu
Tue Jul 27 00:25:11 BST 2004

    [ The following text is in the "Windows-1252" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]


I was trying to figure out the new reservation and back filling feature of SGE 6.0 on a Linux machine cluster.  Apparently, I have difficulty getting that to work.  I have 3 node experimental cluster and am trying to run 3 processor jobs with reservation request.  So far the 2 processor and 1 processor job always finish before the 3 processor job starts.  I was wondering whether I am missing something critical here.  

Below is what I am doing

qsub -N L1 waster1.sh  => 1 processor job
qsub -N L3 -R y -l h_rt=30 waster3.sh  => 3 processor job
qsub -N L1  -l h_rt=100 waster1.sh  => 1 processor job
qsub -N L1  -l h_rt=100 waster1.sh  =>  1 processor job
qsub -N L2  -l h_rt=100 waster2.sh  => 2 processor job

  Still my three processor job is waiting for all the other jobs to finish. I also could not find the $SGE_ROOT/cell/common/schedule file in my installation , which is mentioned in the administration guide. Thanks.

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