Il 19.05.2014 10:01, Arnau Bria ha scritto:
On Fri, 16 May 2014 19:47:51 +0200
Roberto Nunnari wrote:
Hi.
Hi,
Where that nice output comes from? An opensource software? What kind
of monitoring offers on a gridengine cluster? I always use qacct to
see usage SGE stats, or ganglia and big sister for hosts monitoring..
but what could I use for monitoring and drawing graphs about the
usage/load/historic of the queues, jobs, scheduler, etc?
that one comes from a modifcation of:
https://github.com/ganglia/gmetric/blob/master/hpc/sge_jobs/sge_jobs.sh
that generates gmetric (ganglia) graphs.
I did develop a perl dameon that runs in master and generates
ganglia graphs by group (rpm):
https://github.com/pelacables/monpetitami
here are instructions for configuring it and ganglia:
https://github.com/pelacables/monpetitami/tree/master/src
And you also need a modified version of the perl-SGE-API:
https://github.com/pelacables/monpetitami/tree/master/perl-SGE-API
(main modifications are: return hash instead of an array, and job
arrays are not considerd as a single job).
I have some local modifications, if you are interested I could re-pack
it and upload to github.
Thank you and best regards. :-)
Robi
Cheers,
Arnau
Hi.
Thank you very much! :-)
I will give it a try ASAP.
Best regards.
Robi
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
