Minutes of the Computing meeting held on 6 March 2000

Present: G.Corti, G.Gracia, M.Cattaneo,  I.Korolko, P.Binko,   E.van Herwijnen, J.Harvey, G.Wilkinson,
               I.Videau, O.Callot, M.Frank, F.Ranjard, J.Closier, H.Dijkstra, P.Colrain, P.Mato, F.Harris, B.Jost,
               A.Jacholkowska, T.Cass, R.Hierck , F.Loverre, T.Ruf, P.Bartalini, P.Koppenburg, T.Nakada

Minutes : Florence Ranjard

     1.Testing procedure of the split versions SICBMC v231 and SICBDST v231 (F.Ranjard)

        SICBMC v231 and SICBDST v231 are now available for testing.

        The only change compare to v230 released on February 9 is the new pile-up from Andrei which implies new calorimeter
        rawHit banks.

        These versions should be used with dbase v225

To check this new version the following is proposed:

• Agnieszka will run through SICBMC v231 an existing MC tape (~3000 bb-inclusive events) to produce a RawHit tape.
• this RawHit tape will be run through SICBDST v231 to produce DST1 (without pileup)
• this RawHit tape will be run through SICBDST v231 to produce DST2 (with pileup)

• For comparison Agnieszka will run the same MC tape through SICB to produce a DST.

First results with SICBDST v230 :

• missing HCAL banks : I.Korolko is working on the problem
• strange results in the tracking : M.Merk is looking at it
• MUON and VELO seem to work
• As long as the tracking does not work it is difficult to test the RICH.

More results are expected next week with SICBDST v231

Production version : v232

When v231 is certified correct, a production version v232 will be built to contain Agniszka's list of improvements.

The testing procedure should be again applied on v232:

• produce 3000 bb-inclusive MC events with SICBMC
• produce DST1 with SICBDST
• produce DST2 with SICBDST
• run the data quality program on DST1 and DST2

Move SICBDST v232 packages to Brunel v1

when it is done apply the testing procedure on Brunel v1

• produce DST1 with Brunel
• produce DST2 with Brunel
• compare results to DST1 and DST2 produced by SICBDST

The aim is to declare SICB obsolete and to adopt SICBMC-Brunel as production line by end of March.

The situation will be reviewed again at the next computing meeting: Monday 13 March at 14:00

     2. Bookkeeping status (J.Closier)

• the mcupdate and ORACLE databases are now in phase:
  • The ORACLE database is updated every day with mcupdate database.
  • The production updates mcupdate database and a test version of the ORACLE database
• the mcstagein procedure is ready to run on the ORACLE database.
• The mcqry procedure is replaced with a Web interface.
• work to be :
  • review the jobid scheme reservation
  • fill all tables with relevant information in the test ORACLE database
  • trap errors
• Review the status next week.

 

     3. FOCUS report (M.Cattaneo)

           Following points have been raised during the FOCUS meeting held on 2 March 2000:

• W2000 - implementation is going well. Joel Closier is used as guinea-pig
• decomissioning of RISC platforms: proposed calendar
  • no more physics on rsplus end 2000
  • remove rsplus end 2001
  • no cernlib 2001 but a final version cernlib 2002
  • HP and DEC available for LEP experiments only until 2002
• 2nd UNIX platforms: Solaris

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

next meeting : 13 March  2000 at 14:00  
                           bldg 1/1-025