r4 - 24 Mar 2009 - 16:59:46 - KaushikDeYou are here: TWiki >  Admins Web > MinutesDataManageMar24



Minutes of the US ATLAS Data Management meeting, Mar 24, 2009
  • Previous meetings and background : IntegrationProgram
  • Coordinates: Tuesdays, 3:00pm Central
    • (309) 946-5300, Access code: 735188; Dial *6 to mute/un-mute.


  • Meeting attendees: Wei, Armen, Saul, John, Patrick, Wensheng, Hiro, Pedro
  • Apologies: Alexei, Shawn, Charles
  • Guests: None

Topics for this week

  • Storage status/updates - Armen, all
    • Storage status at T1/T2 attached
    • SWT2 waiting for VDT release of Bestman
    • NET2 waiting till corruption problem is solved, Wei will send them version to install
    • Armen discussing with storage team - still testing, but 3 Thorr's will be added soon (90TB)
    • Armen - got 65 TB data on MCDISK last week, adding continuously
    • In principle another 2 Thorr's should be ready soon, 1 being used for testing
  • BNL staging - Hiro
    • Average queued requests past week - not always full, right now nothing queued (Hiro), ~1000 queued (Pedro)
    • Average throughout during past week - 10 TB saturday, 6.7TB sunday, 5TB monday
    • Plans to implement priority and callback
      • Pedro - modified dccp (backed by web service) which will allow priority to be specified
      • Will have intelligent algorithm to sort by high priority as well as files on same tape, balancing priority and tape mounts
      • Web service is capable of callbacks, but still keep polling (perhaps use both, with increased time between polls)
      • This service will also improve pandamover monitoring
    • Average pnfs load past week - was high last week, now stabilized (in fact very low)
      • BNL will keep eye during reprocessing and let Panda team know if it becomes too high
    • HPSS monitoring (Pedro)
  • DQ2 adler32 plugin - Hiro
    • Code is ready
    • Will test in passive mode for BNLPANDA, starting now
    • For Tier 2 sites, we will need more work (pandamover, srm...)
  • BNL migration to space tokens - Kaushik, Pedro
    • Pedro will let us know when 400 TB is available from Thorr's
  • Hot issues
    • File corruption at NET2
      • Wensheng will help with tests to mimic transfer of large files (problem seems to be with ~2GB files)
  • AOB
    • Next week, meeting will start at noon CDT

-- KaushikDe - 24 Mar 2009

About This Site

Please note that this site is a content mirror of the BNL US ATLAS TWiki. To edit the content of this page, click the Edit this page button at the top of the page and log in with your US ATLAS computing account name and password.


pdf 090324_Storage_Status.pdf (8.0K) | Main.vartap, 24 Mar 2009 - 13:19 |
Powered by TWiki
This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback