LHCb Core Software Meeting

Date and Location

05 Feb 2014
10:30 - 12:00
CERN (2-R-030)


Ben Couturier, Gloria Corti, Illya Shapoval, Joel Closier, #Liang Sun, Marco Cattaneo, Marco Clemencic (minutes), Markus Frank, Rob Lambert (Vidyo), Stefan Lohn



  • Default CMTCONFIG switched to x86_64-slc6-gcc48-opt
  • Discussion with ATLAS to use the CMake configuration framework from Gaudi
  • Deadline for Summer Student Projects is 10th of February

Round Table

Marco Cl.

  • ROOT 6 status
    • beta2 tag postponed on my request (several show stoppers to be fixed first)
  • CMake migration
    • verified Bender build
    • migrated and verified Erasmus and MooreOnline
    • migrated Panoramix (verification pending)
    • Urania still to be migrated
    • Joel: what about DIRAC, LHCbDIRAC etc.
  • Gaudi
    • backward compatibility fix/hack for the Plugin Service (allow use of Reflex component names, hardcoded in some Phys code)
    • few minor fixes in CMake configuration
    • there is no support in CMake for ROOT 6 yet (requires action from Benedikt to be able to support builds on the LCG nightlies)
    • added support for the Vera++ style checker (if available), LBCORE-375

Marco Ca.: We will discuss at PAC the time scale for a release of the stack somewhere in February and another one end of March.
After some discussion it's clear that we cannot pick up ROOT 6 beta2 because of deployment (LCGCMT) issues, so we need to do it end of March (after sorting out the deployment issues).


  • LHCbDirac and Ganga in line with LHCb applications (slc6-gcc48), and all the ingredients are ready to use Python 2.7.


  • Went to the SFT meeting where Federico was presenting the plan of work for Geant 4.
    • reported of a vectorizable CLHEP-compatible random generator which probably will be fed back to CLHEP (and planning for a parallel one)
    • GeantV will be kept as an independent project and merge later, but it mentioned the introduction of a new geometry package (USolid, from AIDA)
  • We will need a discussion on the future of the geometry in LHCb (and Gauss), including the work to be done by Marco Cl.'s technical student.


  • Ariadne
    • removed the automatically generated relation of Moore
    • add two more functionalities to the tool: remove all the relations of a node, and remove completely a node
    • few obsolete scripts were removed, now only one script to use: ariadne

Marco Ca.

  • Preparing the Computing Tasks document. The plan is to have a document describing everything the Computing Group is doing, that can be reviewed every year.


  • New branching conventions now in production.
    • Testing the installation of the nightlies on a daily basis (machine currently
    • yesterday there was an outage of CVMFS that blocked Online and Tier-0. There must some improvement
  • We need to clarify the plan for the software installation.


  • People got interested in vectorization of the code.
  • Can we get a release of Moore with vectorization enabled?
    Ben: we need to validate it in the nightlies.

(...discussion on vectorization, performaces an plans ...)


  • Last week we run MooreOnline at the PIT (some problem with the PIT.
  • Suggestion to produce a (text) file from the timing auditor instead of using the standard output.
  • Preparing a mechanism to be able to dump the transient store before an algorithm and resume the processing from that algorithm in a different job
  • Will need a new nightly slot: lhcb-trigger-prod.

-- MarcoClemencic - 05 Feb 2014

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2014-02-05 - 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 & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback