Attending: Matt, Jens (chair+mins), Daniel, John B, John H, Sam, Govind, Winnie 0. Operational blog posts! No more blog posts :-( No operational issues :-) (thus providing matching parentheses) 1. WLCHEPG: sadly none of the attendees have survived yet, or they are attending HEPiX, so the update will have to be postponed till next week. WLCG link (data session): https://indico.cern.ch/event/555063/sessions/207257/ CHEP link: https://indico.cern.ch/event/505613/ 2. DPM workshop - review of agenda: http://indico.cern.ch/event/559673/ We didn't discuss this today (plenty of time) but the agenda is here and you can have a look! 3. Ends, loose (too many to do so we may just cover the most important/interesting or lowest hanging ones and postpone the rest): * Accounting, * Information systems and ATLAS. * use (and status) of test SEs. * data node writeup and Globus endpoints and endpoint implementations and endpoint implementations testing * DPM as cache testing? * IPv6 and data transfers (and test transfers) * Anything for the cloud-at-Crick workshop? * Science DMZ/JISC campus/eInfra, * T2C testing * Starting with T2C testing, and the Hammercloud work which we were due to have reported about half a year ago, the Hammercloud is now in a better state and might be able to do what it was supposed to have done back then; the issue is with dealing with files that are missing etc. More tests and Alastairian prodding are now needed. Durham has been set up but people are more relaxed at the RUCIO end. The JISC campus infrastructure workshop is today, Lydia should be speaking about her experiences with moving data into GridPP using Globus and FTS. https://www.jisc.ac.uk/events/campus-network-engineering-for-data-intensive-science-workshop-19-oct-2016 On the subject of information systems, here's a link to the July GDB where there was a presentation on the information systems working group and accounting https://indico.cern.ch/event/394784/ See also https://twiki.cern.ch/twiki/bin/view/LCG/NewAccountingPortalTesting The risk is that the wheel will be reinvented and the goalposts moved once again but there is effort in the GLUE working group and participation currently from Maria. Anyone interested in contributing should join the GLUE mailing list. https://redmine.ogf.org/projects/glue-wg/ https://www.ogf.org/mailman/listinfo/glue-wg Anyway there is this CRIC thing which may be new, which CERN says "has been discussed for a long time" [at CERN] but it was mentioned at the July GDB, and (re)presented at the WLCG workshop. ATLAS and CMS would be OK with it whereas LHCb will be using a "light" version. The main thing is to not reinvent the wheel, but we also need to see how much additional work we will need to do to bring the various SEs up to the new ways of doing accounting (?) and publishing (?) (not least the CASTOR instance at RAL which famously has a ticket to implement GLUE 2.0) IPv6 in the plot - as generated by the CERN FTS dashboard, with the Imperial FTS server as 'server', looks pretty good - very good in fact, even if the data is a bit confusing? Some sites are sources only (e.g. Bristol), and if it uses IPv6 at Glasgow then it's odd because the Glasgow test SE should be down...? But it looks nice and green, mostly! 3. AOB Cloud workshop at Crick - reminder of a call for abstracts, ideally something involving Real Work(tm), too. Theme is cloud-and-* (and training?), so cloud-grid would be in scope, probably. hepsysman at QMUL - no indication of themes yet; maybe it's open to suggestions? Daniel Peter Traynor: (19/10/2016 09:59:46) hi https://www.jisc.ac.uk/events/campus-network-engineering-for-data-intensive-science-workshop-19-oct-2016 se01 and se02 are test ses at QM don't forget HEPsysman at QM on the 7th November. https://indico.cern.ch/event/577279/ Paige Winslowe Lacesso: (10:37 AM) Does anyone know if HEPSysMan will be available via Vidyo? Daniel Peter Traynor: (10:38 AM) should be able to have it on vidyo