LHCb Core Software Meeting

Date and Location

15 August 2007
10:30 - 11:15
CERN (2-R-030)

Attendees

Gloria, Hubert, Juan, Marco Cl. (minutes), Markus, Patrick (EVO), Philippe

Subjects

News

Philippe:
  • A new version of GRID middleware is being tested (fixes a regression)
  • Bad news for the production:
    • Reconstruction of B inclusive has to be restarted because a minimum bias production was included in the list
    • Productions that used Boole v12r1 have to be deleted (bug in Boole), other version of Boole are being checked
  • No progress on the tracking of Stripping failures (after a re-run of failed jobs produced more or less the same mortality ratio)
Hubert:
  • LCG 53b is being prepared

Software Releases

Gaudi (Hubert, Marco Cl.)

Hubert:
  • The infrastructure for installation of Python modules and scripts is almost ready (fixed the part for configurables). Some tweaking is still needed for win32.

Markus asked when we will move to ROOT >= 5.15. Reflex (in ROOT <=5.14) is defining the typedef longlong which is also defined by DIM in a different way and this causes compilation problems.

Gauss (Gloria)

  • Found why, if an event is generated as empty, it contains the previous event.
    Added a double protection:
    • clean the buffer in case of empty events
    • skip the simulation step for empty events

DaVinci (Juan)

  • David Hutchcroft spotted a problem with PYTHONPATH in some versions of LHCb.

OnLine (Markus)

  • Preparing a new release (not just a re-tag)
    • Need to add Rich Monitoring code, which depends on Lbcom
    Gloria
    it should go in Orwell.

Panoramix (Thomas)

Release and Deployment (Hubert)

  • Fixed a problem in setenvProject.sh (redirection to /dev/null)

Round Table

Marco Cl.
Scalability tests for the conditions database started at CNAF. Few problems have been encountered (after few modest runs, the server hat to be restarted). Investigating.
Hubert
CMT allows to to define a "container" package for a project (in the project.cmt file). It is meant to identify the *Sys package of the project (in LHCb jargon). He will add the correct line to all the projects. (Anyway CMT is not using that parameter in any place)
Markus
Which is the plan to access the CondDB in the OnLine farm?
Marco Cl.: We will produce SQLite snapshots of the database that will be copied locally (shared memory directory, /dev/shm) to the nodes from NFS only when needed.

-- MarcoClemencic - 15 Aug 2007

Edit | Attach | Watch | Print version | History: r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r1 - 2007-08-15 - 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