Deployment team

Tuesday 30 Oct 2007, 11:00 → 12:15 Europe/Zurich

EVO - GridPP Deployment team meeting

Jeremy Coles

- This is the weekly DTEAM meeting - This meeting uses EVO: http://evo.caltech.edu/evoGate/. Join the meeting in the GridPP Community area. - There is a phone bridge available for this meeting. Please call +41 22 76 71400 and use conference ID: 124197. The phone bridge code is 4880.

Minutes dteam-minutes-20071029.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

Ops meeting update ************************* * UK-Ireland / SouthWesternEurope are on CIC duty * The unused and non working service: https://lcg-job-monitor.cern.ch:8443/job-monitor/job-monitor.cgi will be removed later this week. * A new null release was made: glite 3.1 Update 5. This removed some 3rd parties from the gLite externals that were available from DAG. * Oliver Keeble gave an update on SA3 to give an update on progress of service nodes to SL4. http://grid-deployment.web.cern.ch/grid-deployment//cgi-bin/reports.cgi?action=package https://twiki.cern.ch/twiki/bin/view/EGEE/Glite31NodeTracker. The 64-bit SL4 WN has run into a few complications with the desire to include 32-bit libraries in the same release (since a lot of code will still run in 32-bit mode). * SLC3 support will end this week. CERN will deal with its own patches until the gLite nodes are all on SL4. 2 UK sites are still running SLC3. * LHCb has some problems at CNAF last week due to the shared area not working. The problem was related to the migration of the shared areas to GPFS. They suggest "that any important changes in site configuration should be always broadcast at a high level". * No UKI feedback yet on the SLD document circulated last week. Ticket status *************** - See attachment

more information report_file28123.html

11:35 → 11:45 Working towards a position on pilot jobs & glexec

- Summary of discussions to date - Statements of position... this is what we had after the DB in London: "The UK(I) does not accept glexec in its present state and will not recommend its deployment (note to UK only, some sites may ignore that advice). Before we will reconsider its deployment we require to see: a) a satisfactory code inspection wrt security. b) an AUP with VOs agreed in principle by sites (in general through GDB) and all 4 LHC VOs. c) a satisfactory analysis of the effect on the supported batch systems and accounting of identity changes. d) glexec should be simple and safe to configure. Simple so that all sysadmins can understand what they are doing and safe so that misconfigurations do not inadvertently lead to security exposures. e) The three flavours of glexec should exist as separate binaries and not as one with a configuration switch. Each of these should satisfy all package dependencies. e.g., dependencies should express some requirement for an abstract "glexec" property, which is satisfied by "glexec-null" or "glexec-suid" etc. "

11:45 → 11:55 Process to discount SAM instabilities

- Some sites are pre-adopters and ease the migration to new middleware for other sites. We need to have a process that sites can follow whereby periods of SAM instability can be removed from their record.

11:55 → 12:05 Actions review

12:05 → 12:10 AOB

- WLCG reliability workshop http://indico.cern.ch/conferenceTimeTable.py?confId=20080&showDate=all&showSession=all&detailLevel=contribution&viewMode=parallel