LHCb Core Software Meeting

Date and Location

10:30 - 1120
CERN (2-R-030)


Ben Couturier, Gaylord Cherencey, Gloria Corti, Illya Shapoval, Marco Cattaneo, Marco Clemencic (minutes), Markus Frank, Stefan Lohn



  • New version of XRootd (3.2.2), requested by Atlas

Round Table

Marco Ca.

  • Agreed at last PAC: one last Brunel before reprocessing//stripping
    • mainly fixes to reduce memory footprint in stripping (including freeing Boost pool allocators, automatic in LbAppInit)
  • Rob has been working to understand why the FSR merging takes so much memory (it seems that the problem has always been there)
    • Core Software should start helping to tackle the problem (Markus, Marco Cl....)

Marco Cl.

  • Patch from Nathalie to fix issues in GaudiMP (patch #5471).
  • Contacted Coverity support to solve the problem in the upgrade of our installation.
  • Problem with the release of Gauss v41r4 on x86_64: incorrect GAUSS_merged_confDb.py. Shell we re-build as a patch version? NO: just fix release and tarball.
  • Working on Eclipse plugins:
    • added copyright notice and licence
    • minor fixes
    • planning to improve docs and templates
  • CERN official recommendations on copyright and licensing are available:
    • use GPL 3 whenever possible and add the copyright informations to the files
    • shall we adapt Gaudi? the templates? -- send a mail for Gaudi and add to templates
    • to be noticed that we can use public code hosting services, like github.com or gitorius.org.


  • Nightlies are getting out of control: they are too slow.
    Marco Ca.: we need more machines


  • Upgrade patches have been released.
  • Ported GaudiParallelizer to Intel TBB (for the concurrency project), may be try with OpenMP too.
    • Moved the code to the new Gaudi repository.

Marco Cl.: in the context of the Concurrency Framework project, we created a new Gaudi GIT repository for prototyping (GaudiMC.git).

Marco Ca.: Nathalie had problems with comparisons. Marco Cl.: now she is working on late forking, once her patch is applied, we can test/validate using the nightlies.


  • Working on packaging and installation
    • problems with YUM: reimplemented an equivalent tool, but too slow on parsing XML; work a bit more to see if it makes sense to continue
    • testing packaging with CMake


  • Analized log files to estimate the optimal parameters for proxy servers to connect to Frontier/Oracle.

Marco Ca.: what are the outcomes?
Marco Cl.: we need to check the data, but we have good results with a latency in the cache of ~2-3h and the grouping of the queries such that one query can be shared by ~100 jobs/files. After the check, I will contact CMS to see what can be organized.


  • Working on profiling an application so that we can compare the performances with other versions. Problem with local builds.


  • Bunch of releases of Online. The latest one used in the PIT.
  • Problem with the latest GaudiObjDesc dependency on LHCbKernel. Marco Cl. will fix it ASAP.
  • Problem with some change in exception messages. To be investigated.


  • Investigate the memory explosion in FSR merging Markus, Marco Cl.
  • Fix the dependency of GaudiObjDesc on LHCbKernel Marco Cl.
  • Circulate a proposal for adding the license to Gaudi Marco Cl.
  • Organize the tests for the use of Frontier in production.

-- MarcoClemencic - 25-Jul-2012

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2012-07-25 - MarcoClemencic
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    LHCb All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2021 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback