From: Brian Bockelman <bbockelm@xxxxxxxxxxx> Date: 03/07/2016 10:01 AM
> The schedd, however, is more sensitive to number
of queued jobs. Each
> takes up somewhere in the neighborhood of 50KB of RAM and a small
amount
> of CPU time (which becomes noticeable once there are tens of thousands
of
> idle jobs).
I can say from personal experience that there has
been massive improvement in this respect in recent versions, from later 8.2
into 8.4. I've had situations where there's been hundreds of thousands
of jobs queued in a given scheduler, and the only real problem we ran
into was condor_q's proclivity to list every last one of those jobs. Had
to bump up the the condor_q's timeout value for a while before the improvements
came out.