For LHCb, CNAF is currently unusable, so will be dropped from the FEST09 exercise.
Attendance
Please remember to connect to the web conference even if you dial-in, so that we don't have to guess who's out there!
e.g. sip:anonymous is not particularly helpful!
EGEE
Asia Pacific ROC: absent
Central Europe ROC: Malgorzata Krakowian
OCC / CERN ROC: Maite Barroso, John Shade, Antonio Retico
BDII issues raised by three ROCs via CODs. Will be discussed later on.
ENEA-INFO has sprung into life so no escalation is needed. Gareth will reset to 1st step.
Alessandro from Italy had nothing to add for grid operator handover.
Antonio hadn't arrived yet to give his gLite release news, so we jumped to issues from ROC reports, namely the BDII update 34 problems. Laurence had nothing to add. Yet again, the conference system played up, so we rebooted the phone & tried again. These recurring phone problems are a big waste of everyone's time!
EGEE Items From ROC Reports
BDII: new release is in certification (to be installed at your own risk). Antonio said plan for release was to have CNAF, GRIF & AEGIS to keep testing it. Kostas wanted a release date. Antonio: certification this week, wait another week in Production, and then release.
Another issue from Italy with thousands of defunct Globus-gma jobs. Di Qing has released a patch to clean up defunct processes (done this afternoon). Link is in GGUS:42981 for those who want it quickly (patch is being certified). Italy will try out the patch and give feedback.
APEL accounting data to GOC was broken. Spotted by SAM tests and solved. The "too long time to solve" was due to a lack of updates to GGUS ticket by different support entities.
Maite read from the agenda about the complaint concerning the stability of the BDII (ever-increasing functionality but no improved reliability). Laurence commented that the Savannah bug was not about BDII stability, but rather robustness of the Information System (i.e. GFAL failover to different BDIIs). Rémi stated that GFAL (as of GFAL 1.1.0.6) already has BDII failover mechanism for when BDII is down, but that empty responses (i.e. comatose BDII) were harder to handle.
PPS Reports and Issues
As usual, please refer to the link in the agenda for the detailed information!
3.1.100 WMS release is about to be certified. Pilot service will be set up, so Antonio requests a Prod site to deploy the service as soon as the patch is certified, and have experiments use it. Volunteers to contact Antonio.
CREAM pilot has yet another new version, available in repository, for sites participating.
New version of VOMS client with bug-fixes is being certified
Antonio anticipates a (phased) move to production of all the above sometime in February.
gLite Release News
As usual, please refer to the link in the agenda for the detailed information!
gLite 3.1 update 39 in PPS. New version 1.6.1 of VDT available. Used by everyone, but a rollback procedure needed before a roll-out can be considered.
WLCG Items
WLCG issues coming from ROC reports
None
Upcoming WLCG Service Interventions
Consult links on the agenda page.
WLCG Service Coordination
Harry mentioned that the "ATLAS 10 million file test" is now scheduled to run until Sunday (there were CERN FTS problems due to the non-renewal of a proxy on an FTS server).
ATLAS Service
ALICE Service
No report and no one present, but Harry mentioned that they’re setting up a WMS in GRIF this week.
CMS Service
CMS report is on the agenda page.
LHCb Service
Roberto delved into the long-standing issue that LHCb has with the CNAF shared area:
increasing timeout for setting up environment still doesn’t solve CNAF’s shared area problem. (CNAF will buy new per-VO hardware in 2 months, but not fast enough for LHCb). CNAF currently unusable, so will be dropped from FEST09 exercise.
CNAF GGUS:44729 ticket opened prior to Christmas, but closed/re-opened/closed/re-opened several times. Complaint about closing tickets without end-user approval!
native 64-bit O/S still had 32-bit library problems at PIC & CNAF. Extra libraries were needed, so this has been added explicitly to the LHCb VO card. The same problem of missing packages is present in SL5
EGEE Web>SA1>WlcgOsgEgeeOpsMeetingMinutes>WlcgOsgEgeeOpsMinutes2009x01x19 (2009-01-20, JohnShade)
EditAttachPDF
Copyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.