Hey everyone. I'm getting some weird behavior out of my torque/maui
setup, and I thought I'd see if anyone out there can help. Basically,
torque isn't making full use of open resources when I submit large
numbers of jobs via a script.
Our cluster is pretty simple. We have 10 processors, 2 each on three
nodes (cold1, cold2, and cold3) and 4 on cold4. My nodes file looks
like this:
cold1:ts np=2
cold2:ts np=2
cold3:ts np=2
cold4:ts np=4
and maui.cfg looks like this:
QUEUETIMEWEIGHT 0
USERWEIGHT 0
BACKFILLPOLICY FIRSTFIT
RESERVATIONPOLICY NEVER
NODEALLOCATIONPOLICY PRIORITY
NODECFG[DEFAULT] PRIORITYF='CPROCS-JOBCOUNT'
USERCFG[DEFAULT] MAXPROC=10
Things are set up so that each of the 10 processors should get 1 job.
When I submit 10 jobs, things get distributed as I want and would expect
and all 10 jobs run. But if I submit many jobs with a script, like 100,
only 6 jobs at a time run no matter what. What's going on?
Thanks in advance for any help you can provide...Oh, and I'm running
maui-3.2.6p14 and torque-2.0.0p7...
Michelangelo
_______________________________________________
mauiusers mailing list
[email protected]
http://www.supercluster.org/mailman/listinfo/mauiusers
