Antonio is on holiday, but Nick reports that a fix is coming out to prevent ATLAS s/w from breaking WNs (Python version problem w/ logging module). Release will be today or tomorrow to Production. Nick was not aware of anything else coming out for Production.
DECH reported the aforementioned ATLAS problem with Python version (GGUS:50148). It affects many sites!
Germany may impose a ban on the Iranian CA. It is a hot topic in Germany, and France also has the issue. Rolf explained that there are technical solutions - namely blocking the Iranian CA on CEs, in which case it's irrelevant whether it's present on the Worker Nodes.
John explained that SAM checks that all the CAs in the set are present, and that checking for a subset would only work if it were an known entity (e.g. published in the BDII). Although EGEE considers the IGTF approval of CAs as binding, Rolf suggested that this would not be the case in EGI, where there will be an additional “political” selection according to UN and EU rules. Blacklisting is a tricky subject. Rolf said he thought that Christophe Witzig (Switch) was working on an Argus service (central point for banning). Maarten Litmath would know more, as he invented the name
Grid Service Interventions
Nick had spotted an extensive list in GOCDB, and suggested that people check for themselves if so inclined. The link is on the agenda page.
Miscellaneous
gLite 3.0 VOMS servers. Nine servers need to be updated urgently. New, incompatible VOMS client is about to be certified. LAL didn’t have a Quattor profile but it should be fixed for next week. Nick volunteered to get in touch with CERN VOMS support to see if Quattor templates can be shared. Vera: NDGF will also do this next week after holiday. Sven reported VOMS update problems because they tried updating from a very old version (wiped out the database). He suggested warnings in the Wiki.
NA4 made a recent broadcast about Matlab. It is available for EGEE users at very competitive rates (see agenda for links). Contact Vangelis Floros if interested.
Zeeshan was worried that the recent APEL problems mean that accounting data for June is very incomplete. He also pointed out that the T2 Estonia (Baltic Grid) accounting system is broken (missing many hours). Nick was going to contact APEL support to get an update, but the minute taker can provide the following summary:
The initial problem (beginning July) was that data were published in APEL but not transiting to the central repository.
That was followed by a stability problem with the publication mechanism, preventing a lot of sites to publish. SAM correctly reported warnings about last publication date because all these sites were unable to publish correctly.
Latest status (15/7/09) is that "Most problems are now solved on APEL side and publication is working properly at the moment.". The SAM tests show:
Maria was not present for OSG, and the agenda section was from last week. However, Kyle confirmed that the 1st item was taken care of. Not sure about 2nd item. 3rd item – ticket is open, no response yet from administrators.
Review of Action Items
There was one outstanding on Steve Traylen about what sites that only provide storage should do in terms of publishing installed capacity. However, Steve is on a well-deserved holiday at the moment.
EGEE Web>SA1>WlcgOsgEgeeOpsMeetingMinutes>WlcgOsgEgeeOpsMinutes2009x07x13 (2009-08-25, JohnShade)
EditAttachPDF
Copyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.