r3 - 24 Nov 2009 - 14:03:15 - KaushikDeYou are here: TWiki >  Admins Web > MinutesDataManageNov24



Minutes of the US ATLAS Data Management meeting, Nov. 24, 2009
  • Previous meetings and background : IntegrationProgram
  • Coordinates: Tuesdays, Noon Central
    • (605) 715-4900, Access code: 735188; Dial *6 to mute/un-mute.


  • Meeting attendees: Charles, Armen, Bob, John, Hiro, Wei, Shawn, Rob, Patrick, Saul, Wensheng, Michael
  • Apologies: Pedro
  • Guests:

Topics for this week

  • Data deletion - all
    • Continuing ACL issues - BNL, MWT2, AGLT2??
    • Charles - quick check shows directory ACL's were fine, but some file ACL's were wrong (old files, about 3k out of 3M)
    • Hiro - most problems are due to old pilot problem, and also since US pilots are using US production role (need to check if this is still necessary)
    • Plan of action - instructions by Charles and Hiro to check and fix LFC permissions?
    • USERDISK cleanup - Hiro
      • Running at all 5 Tier 2's, but could be having problems at AGLT2 (Chimera?)
      • Deleting records in LFC first, then delete physical files using bulk deletion (50 files)
    • To remind cleaning policy in the US:
      • MCDISK/DATADISK - cleaned by Central Deletion (but may also be cleaned by site only if there is space shortage)
      • MCTAPE/DATATAPE - cleaned by Central Deletion, with notification to BNL to make sure files are fully removed from dcache
      • SCRATCHDISK - cleaned by Central Deletion
      • GROUPDISK - cleaned by Central Deletion
      • HOTDISK - never cleaned?
      • PRODDISK - cleaned by site
      • USERDISK - cleaned by Hiro
      • LOCALGROUPDISK - cleaned by Hiro
  • Tier 3 migration to BNL LFC - Hiro
    • All sites done
  • Tier 2 migration to BNL SS - Hiro
    • Postponed till next meeting
  • Available space status - Armen
    • Lots of data09_1beam, data09_900GeV coming to sites
    • Tier 1 tokens - looks fine
    • Tier 2 tokens - looks fine, lowest available is 40 TB on DATADISK
    • BNL SCRATCHDISK issue from last week, LFC cleaned but storage not cleaned (10TB) - Armen will follow up by email
    • MCDISK plans - expect 100 TB in next 6 weeks (but some already done), and 100 TB early 2010
  • Hot issues
    • dcache instabilities at AGLT2?
      • Shawn - being monitored. If caused by deletion, will find out soon.
    • workernode space issues at NET2?
      • Analysis jobs failing, about 100, running out of space, by Johannes (HC)
  • AOB
    • Hiro - NET2 needs to change srm version to 2.2.0

-- KaushikDe - 24 Nov 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.


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