Deployment team

Tuesday 27 Nov 2007, 11:00 → 12:30 Europe/Zurich

EVO - GridPP Deployment team meeting

Jeremy Coles

- This is the weekly DTEAM meeting - The intention is to run the meeting in EVO: http://evo.caltech.edu/evoGate/. Join the meeting in the GridPP Community area

Minutes dteam-07-11-27.txt

11:00 → 11:20 Experiment problems/issues

Review of weekly issues by experiment/VO - LHCb - CMS - ATLAS - Other

11:20 → 11:35 ROC update

ROC manager update ************************* ROC manager meeting cancelled. Ops meeting update ************************* Ops meeting cancelled due to service reliability workshop at CERN UKI site issues **************** UCL-HEP: Discovered that machine running the CE and BDII_site had load greater than 20. This caused several BDIi drop-outs. Identified the gridice daemon as the culprit, with a process using up to over 50% CPU at times. Had to turn that process off to re-satablish stable functionality. Despite atlas queue being stuffed, we still receive a steady submission fo Atlas jobs. Now queued jobs is close to 1000, with a steadly increasing waiting time (currently at 326.6 Ms). This will inevitably lead to large number of failures due to proxy expiring. Not sure if we should cap the number of queued jobs per queue Monitoring & accounting questions ************************************** Recent SAM problems: Imperial HEP and Cambridge ATLAS (SL tests) problems: Many sites but Durham, Glasgow and Manchester stand out APEL: Problems still seen at: RAL-LCG2 QMUL RHUL UCL-CENTRAL Manchester Durham Bristol RALPP UKI tickets *************** See attached update

more information report_file14456.html

11:35 → 11:40 ATLAS jobs

- Follow on discussion after Monday's email exchanges. -- Changes moving to the panda framework -- Do sites need to be given more information (on Thursday perhaps?)

11:40 → 11:45 Gridmon and use of the WAN

- Update from Barney - Chance to raise/discuss any issues related to networking

11:45 → 11:50 Resources declared in quarterly reports

- We need agreement on what is included ("available" is still not fully defined - Clarification on the instances raised by Duncan (mainly SouthGrid) - First thoughts on special cases for T2 money (£100k being made available for sites with a special case after the formulaic allocations are made and Tier-2s reallocate internally- see mail to T2B).

11:50 → 11:55 Use of and support for VOs

- The RSA work is being proposed to run under geant4/gear - Which sites will support gear and will any stop supporting geant4 if they widen their remit (the geant4 VO manager has now agreed to use of the VO) - It would be good to be supportive of use which follows the correct procedures - What else do we expect (e.g. broadcast, ROC manager or VO manager announcement, formal change to VO scope...)

11:55 → 12:05 Actions review

- Review of open actions here: http://www.gridpp.ac.uk/wiki/Deployment_Team_Action_items

12:05 → 12:10 AOB

- Topics for the UKI meeting on Thursday - Any items for the next GDB? http://indico.cern.ch/conferenceDisplay.py?confId=8508 - ATLAS Tier-1 jamboree 6th December: http://indico.cern.ch/conferenceDisplay.py?confId=23620 - CMS Tier-1 visit (next Friday 7th December)