FacilityWGAPMinutesJun9

Introduction

• FacilityWGAP, IntegrationProgram
• Coordinates: Bi-weekly on Tuesday, 10:30 am Eastern
• ESNET Adhoc, 88-USATLAS, or 510-883-7860, 888728527#
  • See also: http://www.usatlas.bnl.gov/computing/grid/#weekly-meetings

Links

• http://panda.cern.ch:25880/server/pandamon/query?dash=analysis
• hn-atlas-dist-analysis-help - for reference
• http://atlddm02.cern.ch/dq2/accounting/cloud_view/USASITES/30/
• AnalysisQueueJobTests, https://twiki.cern.ch/twiki/bin/view/Atlas/StressTestJob
• AnalysisSiteCertification
• UsTierStressTest - Akira's page
• https://twiki.cern.ch/twiki/bin/view/Atlas/Step09 - Graeme's page
• Stress test jobs on Panda monitor: http://panda.cern.ch:25880/server/pandamon/query?dash=analysis&processingType=stresstest&reload=yes
• HammerCloud jobs on Panda monitor: http://panda.cern.ch:25880/server/pandamon/query?dash=analysis&processingType=hammercloud&reload=yes
• https://groups.cern.ch/group/atlas-dist-analysis-stress-testing-coord/default.aspx
• https://groups.cern.ch/group/atlas-dist-analysis-stress-testing/default.aspx - Tier 2 admins should register here - for Hammer Cloud problems
• https://www.usatlas.bnl.gov/HyperNews/racf/get/StressTest.html

Attending

• Meeting attendees: Rob, Bob, Saul, Patrick, Saul, Tom
• Apologies: none

Status of stress test jobs in US

• last week
  • Issue for failing DB access job at SLAC and SWT2 is now understood. Input files were being copied correctly however were failing in reading. Tadashi found the problem, PyUtils? /AthFile.py failed to check the file, this supports os and rfio only, but needs to support root:// (and dcap://). Sebastien Binet provided a new tag, PyUtils? -00-06-17. Nurcan tried to check out Tools/PyUtils-00-06-17 package locally from SVN so that it will be compiled on WNs with pathena. However currently this package does not compile locally. Investigating.

• this week
  • AGLT2 studies
  • finding lots of jobs sitting in IO_WAIT
  • dccp used as copy input tool at AGLT2, MWT2
  • Bob notes pilots seem to be using gt2
  • Problematic jobs using very little nice cpu.
  • finding disk contention for 8 core machines - maxing out the number of IOPs
  • Need to be able to find out which jobs are running on which nodes
  • Saul: can determine nodes from job types - send list
  • Saul - ANALY_NET2 doesn't see anything in IO_WAIT
  • Patrick - working on direct I/O from xrootd issues
  • I/O - no sustained WAN traffic.

-- RobertGardner - 08 Jun 2009