WARNING: This web is not used anymore. Please use PDBService.MinuteS17January08 instead!
 

Minutes 17 January 08

  • Phone:
    • TRIUMF: Andrew
    • RAL: Carmine
    • NDGF: Olli
    • BNL: Carlos and John

  • CERN:
    • 3D: Dirk, Eva, Andrea
    • ATLAS: Gancho, Florbela

  • Apologies: GridKa, CNAF, SARA, PIC

  • Next workshop: beginning of March – at PIC ? Topics are welcome.
    • Sessions and processes.
    • Move to 64 bits.
    • CCRC in february

  • Sites status:
    • Oracle has asked to install a diagnostic patch and reproduce the ORA-600 problem when dropping the propagation job.
    • Patch fixing the CLOBs problems in LHCb-COOL has been installed.
    • ATLAS Streams setup is hitting new Streams bug related to the use of compressed tables. Logminer is not able to read the redo log / archived log files which contain redo information about compressed tables. Service Request raised to Oracle in severity 1, waiting for fix patch.
    • RAL: Memory leak, querying gv$ dictionary views. Oracle support feedback, solution to be applied together with Critical patch (probably next week).
      • 3rd node for ATLAS RAC coming. New database for TAGs test ongoing (single instance), no Streams monitoring needed, OEM can be setup if needed.
    • NDGF: database was not reachable. Database rebooted. No news about the new cluster availability.
    • TRIUMF: Power outage during Xmas day (EGEE announcement, no 3d list announcement).
      • Database intervention last week?? Failed (multipath....) - plan again for next Wednesday
    • BNL: No problems during Xmas.
      • 32 to 64 bits migration in the same hardware – draft prepared. Carlos will send it around for comments. No testbed to test. Carlos will send the plan and we will check if we can allocate a test system here at CERN.

  • Oracle client sessions and processes:
    • Reconstruction test ongoing on the conditions database, running jobs through Grid. Number of connection was increasing, Carlos increase the number of sessions.
      • Replication affected due to the maximum number of sessions exceeded.
      • Profile used number of session per user higher than the number of processes in the database.
      • After fixing it, replication went well.
      • Also, idle connexions observed in the database.
      • Snipe script implemented. Maximum idle time=1 hour.
    • Request ATLAS to send an email before starting tests like this to the 3d list so sites are aware.
    • dCache: reads calibration data through POOL files, and connexion can be run for long....
    • Global value for the machine set to the maximum allowed. Number for users can be controlled by profiles.
    • SGA: 2G, 8G total memory, 32bits
    • Triumf has implemented the same recommendations. Carlos and Andrew will write a report with the conclusions.

Edit | Attach | Watch | Print version | History: r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r1 - 2008-02-13 - EvaDafonte
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    PSSGroup All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback