Hi Alex, For everyone not interested in Ganga development feel free to ignore but I'm cc-ing the list to keep track of things. I'm preparing a short PR which functionally moves the needed code from GangaGaudi -> GangaDirac and keeps a stub in GangaGaudi to import the methods for backwards compatibility with old (unsupported) LHCb applications. Since GaudiExec (cmake support) has been developed for LHCb the old style LHCb apps are only kept around for backwards compatibility, and I think these are the main users of GangaGaudi who still use upstream ganga. I think it's fair to request a user have a minimal .gangarc file of say: ``` [defaults_DiracProxy] group=gridpp ``` But then with this and running /cvmfs/ganga.cern.ch/runGanga-dirac.sh I would expect they be able to submit a simple executable job which they can't because the executable is pulling in the JDL scripts from GangaGaudi which isn't loaded by default as the code here is largely unneeded and potentially at least pollutes the GPI. I'm re-factoring the config code again to add legacy support for the runGanga-dirac config which should have been updated and this will be in a PR where a user can just submit a simple 'hello world' job against the imperial Dirac server out of the box again. Rob On 09/11/2017 11:36 AM, Alexander Richards wrote:
Hi Rob,
Is GangaGaudi uses in LHCb now or do I understand you that it's now not needed? This bug I also saw and fixed for the vanilla_dirac_fix branch but that branch cant be merged yet as it somehow breaks LHCb.
Essentially the fix is to move the imports from GangaGaudi/Lib/__init__.py back to GangaGaudi/__init__.py where they used to be. Do you remember why they were moved in the first place given that GangaDirac and GangaLHCb all have the imports in the package_root/__init__.py file?
Cheers Alex
On 09/11/17 11:22, Robert Currie wrote:
Hi Daniel,
Can you share your .gangarc and the script/commands which are failing for you, thanks :)
This is a bug in ganga trying to load (effectively LHCb/Gaudi) code from the vanilla Dirac plugin and should be fixed.
(This code should either be in the Dirac plugin if it's needed there or removed as I suspect GangaGaudi is no longer strictly needed/supported after recent developments within LHCb)
As a short term I suspect this can be fixed by either loading the GangaGaudi in your .gangarc, let me know if you need help with that,
Regards,
Rob
On 09/11/2017 08:56 AM, Daniela Bauer wrote:
Hi Daniel,
Possibly related to this ? https://github.com/ganga-devs/ganga/issues/1115
I think this is a ganga rather than dirac issue, but I think the people involved are on this list and will hopefully comment.
Regards, Daniela
On 9 November 2017 at 08:47, Daniel Felea <Daniel.Felea@cern.ch> wrote:
Hello All,
I have got a recent issue when submitting MoEDAL simulations on Grid using Ganga and Dirac. Please see the png file attached, regarding the ConfigError (something with defaults_GaudiXMLSummary). The .gangarc file I have been using is the default one, adapted to VO MoEDAL. And everything goes perfectly when running lxplus or local simulations using Ganga on VM...
I would be more than grateful for any hint on this issue.
Thank you in anticipation !
Cheers, Daniel
-- _______________________________________________ Gridpp-Dirac-Users mailing list Gridpp-Dirac-Users@imperial.ac.uk https://mailman.ic.ac.uk/mailman/listinfo/gridpp-dirac-users [1]
--
Sent from the pit of despair
----------------------------------------------------------- daniela.bauer@imperial.ac.uk HEP Group/Physics Dep Imperial College London, SW7 2BW Tel: +44-(0)20-75947810 http://www.hep.ph.ic.ac.uk/~dbauer/
Links: ------ [1] https://mailman.ic.ac.uk/mailman/listinfo/gridpp-dirac-users
The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
-- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.