Re: [HTCondor-users] PERIODIC_HOLD is applied extremely infrequently

[Vladimir Brik <vladimir.brik@xxxxxxxxxxxxxxxx>]

> I assume that's an OSG metapackage version. What's the output of
> `condor_version`?
$CondorVersion: 8.3.5 Apr 16 2015 BuildID: 315103 $
$CondorPlatform: X86_64-RedHat_6.6 $


> Well that's fun. Do you happen to have any sort of performance
> monitoring on that server, and if so do you see any common patterns
> CPU, memory, etc usage when it works versus doesn't?
The servers are pretty much idle. Most of the time there are at most a 
few hundred jobs running and a few thousand in the queue. Load average 
rarely goes above 0.2.


Vlad



On 05/11/15 15:15, Ben Cotton wrote:
> Vlad,
>
> > > First, what version of HTCondor are you running?
> > 3.3.5
> I assume that's an OSG metapackage version. What's the output of
> `condor_version`?
>
> > > Is the value of "some" consistent (either in raw terms or as a
> > > percentage) across multiple tests?
> >
> > No. Sometimes the right thing happens and all jobs go on hold, sometimes
> > none do.
> Well that's fun. Do you happen to have any sort of performance
> monitoring on that server, and if so do you see any common patterns
> CPU, memory, etc usage when it works versus doesn't?
>
>
> Thanks,
> BC