Re: [HTCondor-users] Core grouping

[Hermann Fuchs <hermann.fuchs@xxxxxxxxxxxxxxxx>]

Hi

The easiest way to make sure a certain software is installed on a
machine is using ClassAds.

On each machine, where you have installed the required software(for
example gnuplot), you define a ClassAd. In your submit file require that
this Class Ad is present.

Example:

On the nodes where gnuplot is installed:
HAS_GNUPLOT = True
STARTD_EXPRS = $(STARTD_EXPRS), HAS_GNUPLOT

In the submit file:
Requirements   = Memory >= 512 && HAS_GNUPLOT =?= True

Then the job will only run on machines which do have the necessary
software.

Best regards,
Hermann

On Thu, 2012-12-20 at 12:22 +0000, Mostafa.B wrote:
> Hi All,
> In our research group, we have a small cluster of 30 cores dedicated
> to the cluster and a few number of temporary cores available in the
> research group network which are added to the cluster whenever they
> are idle.
> Access to the mentioned 30 cores is easy and required third-party
> programs can be installed on their respective PCs without
> difficulties, however the temporary cores are often very hard to
> access.
> Issues arise when a job is sent to one of the cores that doesn't have
> the necessary software to perform the required task. If this was only
> one job, then there would be no major issue. However, when it comes to
> numerous jobs queued, the situation is much more severe.
> In general what I do for running a job in the cluster is, defining the
> executable file as a .bat application which itself runs a few other
> programs that are already installed on the core's respective PC. So
> this clarifies where exactly the problem is happening.
> Has anybody experienced this problem?
> Does anyone know how to tackle this problem without having to remove
> the temporary cores? Probably some way that can specify which group of
> cores may be used for running a specific job.
> 
> _______________________________________________
> 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/

-- 
-------------
DI Hermann Fuchs
Christian Doppler Laboratory for Medical Radiation Research for Radiation Oncology
Department of Radiation Oncology
Medical University Vienna
Währinger Gürtel 18-20
A-1090 Wien

Tel.  + 43 / 1 / 40 400 7271
Mail. hermann.fuchs@xxxxxxxxxxxxxxxx