Hi Steve,
For the VO job count patch, could you please also put on the wiki the
output of your /var/local/condor_jobs_running (or for the pending jobs),
just to make known what is in the file that is parsed by the patch?
I think this could help in the future if sites not running condor need to
set up something similar, and also completes your documentation of the
patch.
Cheers,
Marcus
On Tue, 13 Sep 2016, Stephen Jones wrote:
> Hi All,
>
> After a bit of feedback from some users, various bugs were uncovered in this
> doc:
>
> https://www.gridpp.ac.uk/wiki/Example_Build_of_an_ARC/Condor_Cluster
>
> So I've updated it. Some specific issues were:
>
> a) JURA Archiving into volatile directory (/var/run/urs). It now uses
> /var/urs (thanks Winnie).
>
> b) An improvement to the patch for giving VO job counts in the BDII (thanks
> Marcus).
>
> c) Fix to default_rte_plugin.py that let's it work with xRSL and EMI-ES job
> file inputs. These are sent by ARGO monitoring, and caused chaos for some
> days.
>
> d) Lots of smaller issues cleared up.
>
> The document really needs "a good going over" but I think it's an
> improvement.
>
> Cheers,
>
> Ste
>
>
>
>
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|