Software Meeting

Thursday 10 May 2007, 14:00 → 15:00 Europe/Zurich

32-1-A24

Cristina Biino (INFN Torino)

NA48 Software Meeting

NA48 Software Meeting ============================================================ 10 May 2007 Present: Paolo, Augusto and Cristina + VRVS (Rainer?) I) HEPDB Alan and Ginette had contacts with IT and succeeded in making an update on hepdb in afs. In the future the update will be done interactively by a na48 experts since there is no more a server running all the time as in the past. II) CDR news from Paolo There have been already 4 phone sessions between Venelin and the learning group Simone-Stefano-Emanuele-Paolo. They are changing the scripts on the new CDR machine, at the moment they are working on the watch-dog. A twiki page is available with news and documentation: https://twiki.cern.ch/twiki/bin/view/P326/CDR There will be a meeting with Venelin in Frascati at the end of the month. III) Various news from Venelin and from Jan Van Eldik a) Old CDR PC is kept alive since we do not have yet a completely new working set-up for the CDR. At the moment is still used by users i.e. by the runcompact_pc program. b) Private disk-servers All our private diskservers have a CASTOR backup done by Venelin. These are old disk and they are given back to IT on friday 11 may. They were not used in the last 2 weeks (by the way there was no RFIO access) and only one complaint that has been solved. c) Stage pool The new CASTOR2 stage pool suffered a huge load and this resulted in time-outs for some of the file requests. This happened to .root files read by private programs, and not on compact files, because compact has a scheme to handle time-outs (i.e. tries to open a file, and if not possible waits for some time and then tries again - 5 times). At the same time, to make life more difficult, there was also one of our 5 disk servers not working and various tapes disabled.... The solution was to add 3 additional disk-servers (lxfs6012-6014-6123; recuperated from CASTOR-1) and in that way to distribute the load on more machines. The problem was not with the amount of data (not a disk space problem), but again with the number of files to be staged. - Peoples have practically all stopped using CASTOR-1 (only 3 users left in the last week) ! - Shut-down of CASTOR-1 stagena48 on monday. ----cristina * * * * * * * * * * * * * * * * * Cristina.Biino@cern.ch CERN PH div. tel. 0041.227678729 portable CERN 164792 or 0041.764874792 * * * * * * * * * * * * * * * * *