[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[HTCondor-users] whole machine jobs running with single core jobs

Hi there,
ÂÂÂÂÂÂ I have condor 7.8.5 running on a Rocks cluster. And recently I used the whole machine recipe: https://htcondor-wiki.cs.wisc.edu/index.cgi/wiki?p=WholeMachineSlots.

  Â With that, I can see slot 25 created on each of the 24 cores executing nodes. As far as I can tell the configuration works as expected when condor queue is not totally full. When it is full single core jobs end up running in the same node already occupied by whole machine jobs (slots 1 to 24 are claimed at the same time that slot 25 is). And vice versa (slot 25 is claimed in the same node where single core jobs are running).

ÂÂÂÂÂÂ Shouldn't the machine where slot25 is claimed be blocked from receiving other jobs with the configuration above? Did I misunderstand something?