Pilot of CREAM-CE extended to 16th of December.
GOCDB service name cleanup will take place on Wednesday.
LHCb and ATLAS looking forward to the day that SRMv2 tests will count for availability.
Attendance
EGEE
Asia Pacific ROC: ShuTing Liao
Central Europe ROC: Malgorzata Krakowian
OCC / CERN ROC: Diana Bosio, John Shade, Antonio Retico, Nick Thackray, Steve Traylen
French ROC: 0033478930880 (=David Bouvet + ?)
German/Swiss ROC: Angela Poschlad
Italian ROC: Absent
Northern Europe ROC: Gert Svensson
Russian ROC: Lev Shamardin, Victor Edneral
South East Europe ROC: Ioannis Liabotis
South West Europe ROC: Kai Neuffer
UK/Ireland ROC: Jeremy Coles
GGUS: Torsten Antoni
GOCDB: Gilles Mathieu
WLCG
WLCG Service Cordination: Harry Renshall, Jamie Shiers
Unknown affiliation: catalin Condurache.
=> Please use conference's web interface to enter correct name & affiliation. It makes the minute taker's life somewhat easier, and reduces the risk of missing attendees (e.g. multiple participants in conference rooms).
Feedback on Last Week's Minutes
There were no comments on the minutes which Steve, exceptionally, had read.
EGEE Items
Grid Operator Hand Over on Duty
Primary Team
Secondary Team
From
ROC Russia
ROC UKI
To
ROC CE
ROC AP
Site BEIJING-CNIC-LCG2-IA64 will be suspended immediately after the meeting for having failed to respond in any way to their APEL problem.
Ioannis is waiting for feedback from SE sites. Gert will check the targetted sites of the NE ROC.
David Bouvet asked NE ROC about ITPA-LCG2 which has been targetted for suspension for several weeks. The COD would like an answer after 4 weeks of no responses. Ron will check. David pointed out that the Ops Manual says that COD can suspend site in the case that the ROC doesn’t answer - is this really the case? Steve said this shouldn’t happen (ROC not answering). John pointed out that the ROC's availability will suffer, so it’s in their best interest to answer. David replied that it once took 1.5 months to suspend a Russian site.
In this particular case, the original problem appears in GGUS and in previous Ops meeting minutes. Jeremy suggested that an ACK from the ROC to the original mail would be good (sometimes things get mis-routed). Gert pointed out that the flood of GGUS tickets and e-mails is a problem (there's a current discussion in LCG-ROLLOUT about broadcast spam), but Steve retorted that this is what the agenda & minutes of the weekly Operations Meetings are for! Gert will chase the NE sites.
PPS Issues and Reports
Antonio said that there were no issues, just some tidbits of information:
Minor review of the PPS Service Description document
CERN-PROD (and a few other sites) were bitten by a recent YAIM bug in config_wn on 64-node running 32-bit gLite s/w. A fix is on its way.
Fifth checkpoint of CREAM-CE; CMS is actively using the service. Also some progress to monitor with Nagios. Pilot will be extended to 16th of December.
includes first release of Hydra (pronounced Heedra if you're Greek or Antonio), for which Antonio requested sites to help with future testing (looking at the biomed community).
New version of FTA log-rotate is being fast-tracked to Production, as is a new version of CREAM with fixes LSF interaction issues.
Delay of last Thurdays’s release was due to insertion of FTS fixes and a small problem which needed debugging. Should be out today.
EGEE Items
None from the ROC Reports
For the new GOCDB service names, Gilles said that the agenda was self-explanatory. There was a short discussion about Central-LFC and Local-LFC (assumed to be read-only). Gilles will only update LFC service name changes if things get cleared up quickly.
Diana reported that GOCDB allowed downtimes that started after they'd finished. Gilles remarked that this would be wonderful for increasing site availability, and that his test and validation systems didn't allow it. It turned out that Diana was off by a whole year, which led to the thought that there should be a maximum allowed downtime (a year's downtime is senseless).
WLCG Items
WLCG issues coming from ROC reports
None
Upcoming WLCG Service Interventions
see the usual link on the agenda page.
ATLAS Service
ALICE Service
CMS Service
LHCb Service
The LHCb report "appeared by magic" during the meeting.
Roberto announced some new SAM tests for LHCb. Site calculations had been flawed by LHCb CE tests not working properly, but SE tests were OK. Roberto now using Dirac unit tests plus some tests copied from the ops SAM tests.
LHCb have moved everything (SRM) to SE tests. Alessandro said that ATLAS had done the same thing and asked about SRMv2 tests. John said that tomorrow’s MB will discuss the issue of adding SRMv2 tests to site availability calculations, so stay tuned.
Steve took the opportunity to ask why LHCb tests were failing. It was because of Dirac "not doing things properly".
Roberto gave advance warning to sites that the pilot voms-role configuration request will be broadcast
He also asked about a week-old, non-critical GGUS ticket about stream replication not working with GridKa. Angela admitted being surprised by the "no streams here" answer from the GridKa database people, and will follow-up. Participation in a workshop had prevented her from responding to the ticket earlier.
WLCG Service Coordination
Neither Harry nor Jamie were present, but Steve mentioned that the recent GDB and WLCG workshop had discussed storage topics and that he would provide links.
Rob fumbled the mute button and hung-up by mistake, but re-joined to hear Maria mentioning two tickets. One she had closed as unsolved, but wasn't sure what to do with the other (GGUS:43263) about VOMS server version. She wondered how a French user could make this urgent request to Fermilab whilst the thing was in pre-production. Chairman Steve suggested that the ticket should be closed as unsolved and tracked in Savannah.
Another ticket about ATLAS publishing to top-level BDII should be closed (by policy, ATLAS doesn’t want to do this). Diana was confused about Fermilab being in OSG and the CERN ROC, but it would seem that the site split, and only storage is left in the CERN part. Steve quoted Ian Bird as saying that "if you’re not in the top-level BDII, you’re not in WLCG", but added that Ruth may disagree.
EGEE Web>SA1>WlcgOsgEgeeOpsMeetingMinutes>WlcgOsgEgeeOpsMinutes2008x11x17 (2008-11-19, JohnShade)
EditAttachPDF
Copyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.