Harry: info about upcoming FTS fixes
Antonio(off-line): This update fixes the FTS web service's delegation code that the delegated proxy certificates should
not be corrupted BUG:33641 . The fix is to be applied server-side and requires only a restart of the service after
upgrading the glite-data-transfer-fts package.
EGEE Items From ROC Reports
ROC France:
GRID2-FR certificate were rejected by VOMS servers behind the alias voms.cern.ch. Steve Traylen fixed the problem this morning.
ROC SEE:
Sites in SEE requested to give top priority to bdii version 5 certification. There has been no progress in https://savannah.cern.ch/patch/?2623 for a month. This prevents more top bdiis to be deployed to better load balance as the current glite production bdii has numerous issues (GGUS ticket 43578)
Laurence(off-line):Patch 2623 was rejected as a newer version of the BDII was created due to issues raised by some early adopters. The new patch is 2783 which is currently being tested. PATCH:2784
I am aware of one packaging issue which will result in this patch being rejected but I am waiting for the package to be rejected to see if there are any other issues which will need fixing. As soon as I have this information I will create a new patch.
I notice that PATCH:2784 is current assigned priority normal. This should be increased to high if this patch is seen as important.
Antonio will follow-up with the EMT
Follow-up with EMT the re-prioritisation of PATCH:2784 and possibly increase it to high UPDATE 20-Mar: I discussed with the EMT Coordinator. There is a long list of services which now have priority both in certification and release preparation. They welcomed the idea of a pilot service to be run at some sites, which by providing real-usage records would help them making the certification faster. I am preparing a request for sites to join this activity to be presented at the next OPS meeting.
UKI-LT2-UCL-HEP - New CE now online and passing SAM tests. Decided to shorten downtime in GOCDB. Downtime lifted on 05/03 at 16.05. It took 10 hours for this information to be propagated through the system (e.g. SAM).
John: This is not normal gridview should get in synch with GOCDB every half hour.
Antonio: this is possibly related to the GODB connection problems experienced last Friday and reported via e-mail to the GOCDB discuss list
WLCG Items
WLCG issues coming from ROC reports
NE ROC :
Recently there has been a request from WLCG to reinstall the WNs with 64-bit SL5. In principle this is OK. We have had similar requests also from non-HEP users in the past. However, a substantial fraction of our user community depends on the UIs to compile their codes. So we would very much like the UI to be on the same platform as the WNs. For this reason we would like to have a 64-bit SL5 glite-UI or a 32-bit glite-UI which is certified on SL5 64-bit. Will such glite-UI release be made available and if so, when will it be available?
Oliver (off-line): A gLite3.2/SL5/64 UI will be made available. Indeed, all services will eventually be made available on that platform.
The situation with the UI is that we have 99% of build, but zero runtime experience with a number of components on there (eg LB 2.0).
Andreas Unterkircher will be covering the subject (gLite 3.2 roll-out plan) during Wednesday's GDB (http://indico.cern.ch/conferenceDisplay.py?confId=45473 )
Upcoming WLCG Service Interventions
Consult links on the agenda page.
WLCG Service Coordination
Harry:A request was moved to re-schedule to the 1st of April the intervention on cern networks foreseen for the 18th of March. Apparently this is not possible. Updates to be communicated at this table and at the daily WLCG ops meeting
ATLAS Service
-
ALICE Service
-
CMS Service
-
LHCb Service
Harry: LHCb performed a "GGUS alarm ticket test". This went well and they will run another one this afternoon more reflecting a real LHCb production use case
Roberto: LHCb phoned directly castor service admins at CERN-PROD for an issue observed last Thursday. Basically there is an issue with Globus xio where there are many zombie processes running on the gridftp server which need to be cleaned-up. A GGUS:46946 was opened as well with top priority, the issue was initially fixed by the administrator contacted by phone and now it is happening again preventing LHCb users from working. In all this the GGUS ticket worklog is empty so the ticket seems stuck at the ROC level.
Antonio: the ticket is not stuck because it is being followed-up by the Cern ROC as a standard team ticket (same workflows as for the CODs). The fact that the work is not being logged is disappointing though and will be followed up
Helene: A combined release of CIC Portal and GOCDB with brand new downtime mechanism is in preparation. The new mechanism allows the users to subscribe to notification thus avoiding avoiding "spamming" of broadcast messages. A further reminder will be issued on Monday and unless serious issues are reported the roll-out is scheduled for Wednesday 18th of March. In order to benefit of the new features the users will need to apply configuration changes, otherwise the behaviour of the interface stays unchanged.
Maria (via e-mail): Please record in today's minutes the announcement of the dedicated meeting we shall hold on 12/3 on GGUS-OSG ticket routing for Tier1s. I shall include agenda and notes on the next Ops meet. agenda.
Next Meeting
The next meeting will be Monday, 16 Mar 2009 15:00 UTC (16:00 Swiss local time).
Attendees can join from 14:45 UTC (15:45 Swiss local time) onwards.
The meeting will start promptly at 15:00 UTC (16:00 Swiss local time).
The WLCG section will start at the fixed time of 15:30 UTC (16:30 Swiss local time).