Re: [HTCondor-users] Pre-empting two jobs

[Brian Bockelman <bbockelm@xxxxxxxxxxx>]

Hi Evan,

I assume you are referring to a partitionable slot setup (also called âp-slotsâ)?

In that case, I think you are looking for a feature called âpartitionable slot preemptionâ.  According to the ticketing system, it is slated for the 8.3.6 release (preliminarily targeted for June 18).  That said, I donât see documentation for the feature yet, so I donât know if itâs going to slip.

Brian

> On Jun 16, 2015, at 8:13 PM, Evan Sarmiento <esarmien@xxxxxxxxxxxxx> wrote:
> 
> Hi All,
> 
> I was wondering if there was a way to configure HTCondor to pre-empt two jobs.
> 
> Letâs say that I have a job that requires 4GB and 1 CPU on an HTCondor pool of 4GB,2CPU. Two jobs have consumed the entirety of the resources of the HTCondor pool, each job consuming 2GB of memory and  1 CPU.
> 
> Now, I read all the documentation and apparently the solution here is to use âdefragâ since these will be dynamic slots or compute-on-demand, but, compute-on-demand requires that you actually specify the startd address of the host you want to run the job on.
> 
> Is there any magical HTCondor configuration parameter which will allow a job to pre-empt as many jobs-on-dynamic-slots as it needs to in order to run?
> 
> Best,
> Evan
> 
> 
> Evan Sarmeinto
> Front End Systems Software Developer
> Harvard-MIT Data Center
> T: 617-384-5876
> E: esarmien@xxxxxxxxxxxxx
> Follow us: @hmdcrce
> _______________________________________________
> HTCondor-users mailing list
> To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx with a
> subject: Unsubscribe
> You can also unsubscribe by visiting
> https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users
> 
> The archives can be found at:
> https://lists.cs.wisc.edu/archive/htcondor-users/