Minutes of the Computing meeting held on 17 April 2000

Present: G.Gracia, A.Schopper, A.Jacholkowska, M.Needham, J.Harvey, I.Korolko, K.Harrison, S.Amato,
               F.Ranjard, E.van Herwijnen, T.Nakada, I.Videau, G.Wilkinson, F.Fiedler

Minutes : Florence Ranjard

   1. SICBDST v232 is released

This version will be used by the Calorimeter group to produce DST1 files

• Already 50K DST1 have been produced on Linux..
• If any improvements are required in the Calorimeter code, new versions of corresponding package(s) will be released
  and used in a revised version of SICBDST v232 : SICBDST v232r1,r2,...
• If any bug fixes are required in detdes or/and dbase
  • commit should be done on a branch of detdes v3r2 and/or dbase v226
  • new versions will be used in a revised version of SICBDST v232
  • commit should also be done on the head revision to be used by SICBDST v233,...
• Latest version of calorimeters will be used in SICBDST v233

    2. mcqry and mcunder

The calorimeter group needs to reconstruct old MC files produced with SICB v114/v115. To retrieve files they use mcqry and
mcunder. Unfortunately the ascii data base is corrupted since last week and nobody knows how to restore it.

• Since only old data files are required, Florence will try to find a backup.
  • $LHCBHOME/scripts/mcqry03 and mcunder03 are now accessing mcmain.db.310300

    3. Next versions of SICBMC and SICBDST

• SICBMC v232: forseen improvements:
  • simgeom, simvdet: VELO is in magnetic field.
  • new RICH geometry
  • new magnetic field map : keep the ascii file on all platforms.
  • corrections in simmubg for pileup.
  • dbase will be a package part of the requirements file : no need to specify the dbase version number in the job.
    • time scale: after Easter, A test version should be available for the next meeting (May 8)

• SICBDST v233 : forseen improvements:
  • axselect: new package
  • kalman: magnetic field z<95.
  • digmuon, trimuon : improvements.
  • trivert: improvements
  • new data base package used by SICBDST
    • time scale: after SICBMC v232

     4. A.O.B.

• Check with Joel the staging pool (1 pool or 2 pools?)
• The profiling of SICBMC v231 on NT has been done by Eric:
  • For Bs events: 75% of CPU time is spent in Pythia,
  • For Bd events: 35% of CPU time is spent in Pythia
• Paolo will study the problem when he is back from vacations.

===============================================================

next meeting :  8 May  2000 at 14:00  

                           bldg 1/1-025