Deployment team & sites

Tuesday 18 Jan 2011, 11:00 → 12:16 Europe/London

EVO - GridPP Deployment team & sites meeting

Jeremy Coles

- This is the biweekly DTEAM & sites meeting - The intention is to run the meeting in EVO: http://evo.caltech.edu/evoGate/. Join the meeting in the GridPP Community area. - The phone bridge number is +44 (0)161 306 6802 (CERN number +41 22 76 71400). The phone bridge ID is 77907 with code: 4880.

Minutes DteamMinutes18Jan11

11:00 → 11:20 Experiment problems/issues

Review of weekly issues by experiment/VO - LHCb -- Are the transfer problems back? http://hepwww.rl.ac.uk/nraja/UKUploadProblems/index.html. - CMS - ATLAS -- "ATLAS are in the process of decommissioning of the MCDISK space token . Their aim of unifying DATADISK and MCDISK starts by moving datasets from MCDISK to DATADISK. With this in mind; please can sites in the UK who currently have more than 5TB of free space in their ATLASMCDISK space token please reduce the MCDISK space to a size that leaves the space token with only 5TB of free space. Please allocate this freed space into the ATLASDATADISK space token." - Other -- lcg-cr fails on QMUL. 'srm://se03.esc.qmul.ac.uk//t2k.org/nd280/v8r5p11/unpk/ND280/ND280/00005000_00005999//oa_nd_spl_00005007-0003_ot3a2qrmcuec_unpk_000_v8r5p11.root: Invalid argument\n', 'lcg_cr: Invalid argument\n'] --- "This error (which is horribly non-specific) is an issue with communication with the BDII used to get information about the source and destination systems. It is likely that the issue is not with se03, but with QMUL WNs <-> the RAL and Imperial Top-level BDIIs." "Assuming the jobs specify their data requirements, then they should not be trying to pull data from QMUL." - Experiment blacklisted sites - Experiment known events affecting job slot requirements - Site performance issues

11:20 → 11:35 Meetings & updates

- ROD team status (any points to raise to sites or issues to follow up?) - Tier-1 update Operational security -- Checking results at https://pakiti.egi.eu Site availability & reliability for December -- Let Jeremy know of any serious problems with the figures http://gvdev.cern.ch/GRIDVIEW/downloads/Reports/201012/wlcg/WLCG_Tier2_Dec2010.pdf Summary from GDB last week - Agenda: http://indico.cern.ch/conferenceDisplay.py?confId=118230 - EGI-UF 11-15th April; WLCG Workshop, DESY, Late June/Early July; EGI Technical Forum, Lyon, 19-23rd September. - The 2011 schedule update http://indico.cern.ch/getFile.py/access?contribId=5&resId=2&materialId=slides&confId=115356. - CREAM status reviewed. Experiments generally seem ready so the request is now for every site to deploy a CREAM CE. A date for the phase out of the LCG-CE was not decided but a timeline of "as soon as possible" was mentioned. Some non-LHC VOs still use the LCG-CE. - Most of the meeting was used to hear updates from the Data Access and Management Demonstrators. These are now generally well demonstrated and will be taken up within the experiment deployment structures (except for CoralCDN and Cassandra/Fuse as LFC/SRM alternative which are being dropped).

11:35 → 11:45 Site status for....

CREAM - which sites do not yet have a CREAM CE in production? gLite-APEL - have any sites not yet moved? ARGUS - This is likely to be a theme in the coming months. The MB needs to re-discuss the status of the multi-user pilots policy but expect more activity soon. Do we still have any working deployments!?

11:45 → 11:50 WLCG Tier-2 availability, accounting and misc results

Site availability & reliability for December -- Let Jeremy know of any serious problems with the figures http://gvdev.cern.ch/GRIDVIEW/downloads/Reports/201012/wlcg/WLCG_Tier2_Dec2010.pdf Question for Kashif: Can we establish our own process for modifying erroneous Nagios entries? Accounting now appears up-to-date over the "broker down period". Are there any remaining concerns in this area? The dteam VO VOMS changes. Details are in the ticket here: https://gus.fzk.de/ws/ticket_info.php?ticket=65307. On the last check (12th Jan) these sites had nodes of concern: RAL-LCG2 gridpp.rl.ac.uk RAL-LCG2 rl.ac.uk UKI-LT2-Brunel brunel.ac.uk UKI-LT2-IC-HEP hep.ph.ic.ac.uk UKI-LT2-RHUL ppgrid1.rhul.ac.uk UKI-LT2-UCL-CENTRAL ccc.ucl.ac.uk UKI-LT2-UCL-HEP hep.ucl.ac.uk UKI-NORTHGRID-LANCS-HEP hec.lancs.ac.uk UKI-NORTHGRID-LANCS-HEP lancs.ac.uk UKI-NORTHGRID-LIV-HEP ph.liv.ac.uk UKI-NORTHGRID-MAN-HEP tier2.hep.manchester.ac.uk UKI-NORTHGRID-SHEF-HEP shef.ac.uk UKI-SCOTGRID-DURHAM dur.scotgrid.ac.uk UKI-SCOTGRID-ECDF ecdf.ed.ac.uk UKI-SCOTGRID-ECDF glite.ecdf.ed.ac.uk UKI-SCOTGRID-GLASGOW gla.scotgrid.ac.uk UKI-SOUTHGRID-BHAM-HEP ph.bham.ac.uk UKI-SOUTHGRID-BRIS-HEP phy.bris.ac.uk UKI-SOUTHGRID-CAM-HEP hep.phy.cam.ac.uk UKI-SOUTHGRID-OX-HEP physics.ox.ac.uk UKI-SOUTHGRID-RALPP pp.rl.ac.uk An update came at the EGI ops meeting on Monday: dteam VOMS server change from CERN to GRNET (Dimitris). CERN VOMS server to be shut down next wednesday, 26th Jan; sites should remove CERN settings. -- gstat2 publishing The web view is here: http://gstat-prod.cern.ch/gstat/summary/GRID/GRIDPP/ 1) We notice that site numbers are correct to about the 10% level (often much better). Remember to update what is published here when you add additional resources. 2) Two figures for total storage are seen. One under "summary view" and at the top of the "site view" page and another on the storage plot (middle). Probably a units issue GB vs GiB. 3) Details at the VO level are largely missing. The WLCG MB want this level of installed capacity figure to be able to check against pledges. The numbers are NOT used for any kind of scheduling. The numbers should be taken from the GridPP planning spreadsheets.

11:50 → 12:00 Tier-2 quarterly reports

- Quick review of remaining concerns in the reports - Chance for sites to comment on the information/trends

12:00 → 12:05 AOB