LHCb Core Software Meeting

Date and Location

27 Nov 2013
10:30 - 11:52
CERN (2-R-030)

Attendees

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

Subjects

News

  • LIM: release planning
    • next release will be for validation of ROOT 6 (gcc48 only) against ROOT 5 (gcc46+gcc48, slc6 only)
    • future releases will be only on gcc48

We discussed the internal schedule and deadlines for validation releases.
Marco Ca. reminded that we agreed to get a stack for validation that needs slc5-gcc46, slc6-gcc46 and slc6-gcc48, including the changes by Markus for the federation (GaudiUtils).
We agreed to make a new Gaudi (v24r2) including Markus' changes based on the latest LCGCMT, then Marco Cl. will prepare Gaudi v25r0 with the changes required to compile with both ROOT 5 and 6. In this scenario, we will not need gcc46 builds at all for the ROOT 6 validation, and we can declare that the stack on Gaudi v24 is the last one required to support the old C++ standard.

Round Table

Marco Cl.

  • Nightlies
    • instabilities with the new scheduling system on Tuesday, probably a problem with limits on processes/files
    • much better stability today, after increasing the limits for lhcbsoft
    • as a side effect, we are now testing the new binary tarballs produced in the nightlies (containing only the InstallArea)

Gloria: there is something strange in the Gauss tests (probably due to missing build products in the source directories)

  • ROOT6
    • issues with networked I/O (memory corruption), Axel is investigating
    • EventPacker: the I/O headers must be public for the migration to ROOT 6
  • CVMFS configuration (triggered by Beat's findings)
    • root catalog too big and downloaded every hour: we should use more the hierarchical catalog
    • Jakob can help setting up a vobox (a puppet template should be available) to both distribute the nightly builds and test the version 2.1 of the server

Joel

  • puppet template for the build machines ready, but with problems (git missing)

Markus

  • Suspicion that CVMFS can get corrupted. Jakob is investigating.

Gloria

  • Call for a meeting (Witek) in January to discuss the evolution of HepMC.
    • They would like to have the development done so that the experiments can pick it up for Run2.
  • We need a non-public package for Rivet
  • LPCC Workshop last week of March focusing on detector simulation.

Marco Ca.

  • Meeting/Workshop with OpenLab partners in two weeks time, focused on discussion on the future role of OpenLab
    • send ideas/suggestions to Marco Ca.

Stefan

  • Improving the Jenkins part of LHCbPR
  • Trying to speed up the LHCbPR framework itself
  • Setting up a prototype test-job for Hlt in Jenkins (not integrated with QMTest as Rob would like)

Ben: discussion on the

Ben

  • Had to change the group login for EOS
    • it was setting LD_LIBRARY_PATH breaking xroot (wrong version)
    • now using the new EOS setup script
  • Working on the nightly scripts (allowed to install a project with dependencies)

Rob

  • Migrating stuff to EOS
    • several glitches/annoyances, even in Dirac
  • Reminding the build on gcc43 of an old stack to be able to run L0App
    • agreed to wait for the new Dirac, unless it gets delayed further


-- MarcoClemencic - 27 Nov 2013

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2013-11-27 - 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