Minutes of the GridPP storage EVO session, 15 June 2011. Present: RHUL: Govind Edinburgh: Wahid Glasgow: Sam, David Lancaster: Matt Liverpool: John, Stephen Oxford: Ewan QMUL: Chris Sheffield: Elena Manchester: Alessandra Imperial: Duncan RAL T1: Brian, Jens (chair+mins) 1. WLCG workshop - only Brian is going, Wahid might be going as well but maybe not. No immediate suggestion as to topics to raise, they are still looking for contributors. We might talk about SpeedySRM if there was something new to report? 2. AHM: One NeISS paper went in, with contributions from GridPP (Sam). 3. Small VOs - Brian submitted a quick list replicated here: n #sites total Lancs L'pool Shef Durham Glas IC QMUL RHUL MAN ECDF biomed 8 3245 105 129 211 1000 600 779 290 131 supern. 6 3989 481 126 295 1500 0 250 0 1337 hone 5 553 24 26 0 0 0 43 23 437 mice 2 2289 0 0 0 0 1500 789 0 0 superb 2 14 0 0 0 0 0 0 14 0 pheno 2 80 0 0 0 0 80 0 0 0 fusion 2 18 16 2 0 0 0 0 0 0 neiss 1 2000 0 0 0 0 2000 0 0 0 pheno 1 13000 0 0 0 13000 0 0 0 0 calice 1 2 0 2 0 0 0 0 0 0 esr 1 67 0 67 0 0 0 0 0 0 compch. 1 1294 0 0 0 0 0 0 0 1294 dzero 1 505 0 0 0 0 0 505 0 0 Who should we bug? It makes sense to bug the ones with the most data, e.g. (1) Supernemo, (2) H1, and (3) MICE. MICE are doing MC at Glasgow, otherwise using the grid. Dave Colling also involved with MICE. The idea is we get them into the meeting to talk about their data management, and what their problems are, and we can then tell them about the services we have (e.g. consistency checking of data at various levels, or, to some extent, handholding.) Then we can decide together what to do and how to support them better, if they need that. 4. [10:02:55] Govind Songara joined [10:03:25] David Crooks joined [10:03:25] Matthew Doidge joined [10:03:28] Brian Davies joined [10:03:42] John Bland joined [10:04:17] Stephen Jones joined [10:04:57] Ewan Mac Mahon joined [10:05:06] Ewan Mac Mahon left [10:05:14] Queen Mary, U London London, U.K. joined [10:06:04] Elena Korolkova joined [10:08:43] Brian Davies n # sites total Lancs Liverpool Sheffield durham Glas IC QMUL rhul man ecdf biomed 8 3245 105 129 211 1000 600 779 290 131 supernemo 6 3989 481 126 295 1500 0 250 0 1337 hone 5 553 24 26 0 0 0 43 23 437 mice 2 2289 0 0 0 0 1500 789 0 0 superb 2 14 0 0 0 0 0 0 14 0 pheno 2 80 0 0 0 0 80 0 0 0 fusion 2 18 16 2 0 0 0 0 0 0 neiss 1 2000 0 0 0 0 2000 0 0 0 pheno 1 13000 0 0 0 13000 0 0 0 0 calice 1 2 0 2 0 0 0 0 0 0 esr 1 67 0 67 0 0 0 0 0 0 compchem 1 1294 0 0 0 0 0 0 0 1294 dzero 1 505 0 0 0 0 0 505 0 0 [10:08:46] Alessandra Forti joined [10:10:19] Alessandra Forti it is [10:10:33] Ewan Mac Mahon The grid is like the cloud, but you don't have to install the OS yourself. [10:12:20] Stephen Jones Just because it's called a "cloud" doesn't mean that the definition should be "cloudy". [10:13:12] Ewan Mac Mahon It might be interesting to know how this chap wound up on GridPP resources rather than NGS - they have an irritating habit of hoovering up the small folks. [10:13:25] Jens Jensen [10:13:27] Sam Skipsey Because he knows us personally? [10:13:38] Duncan Rand joined [10:13:48] Alessandra Forti very advance computing concept [10:13:54] Alessandra Forti know the admins [10:15:14] Ewan Mac Mahon Mice seem to be in a constant state of being about to grow into a decent sized VO with serious data. [10:15:53] Sam Skipsey Indeed, Ewan - but as far as we can tell, there's only a few of them who actually do anything on the grid... [10:15:59] Ewan Mac Mahon SNOplus are also 'about to be' doing things, AIUI. [10:16:35] Ewan Mac Mahon Yup. We might just want to get some exploratory feelers out; they certainly don't seem to have a problem to solve yet. [10:16:49] Queen Mary, U London London, U.K. Lustre workshop 14 July QMUL [10:18:33] Brian Davies http://lcgwww.gridpp.rl.ac.uk/cgi-bin/fts-mon/fts-mon.pl?q=transfers_count&p=12hours&v=Group+all+vos&c=Group+all+channels&s=Failed&r=All&i=&.submit=Submit+Query [10:23:37] Brian Davies http://lcgwww.gridpp.rl.ac.uk/cgi-bin/fts-mon/fts-mon.pl?q=transfers_count&p=12hours&v=Group+all+vos&c=All&s=Failed&r=HTTP_TIMEOUT&i=&.submit=Submit+Query [10:25:29] Wahid Bhimji yeah there is transfers_count but no channels_count [10:27:15] Wahid Bhimji That is useful monitoring ... [10:27:53] Alessandra Forti yeah [10:28:21] Elena Korolkova left [10:28:21] Brian Davies left [10:28:21] David Crooks left [10:28:22] John Bland left [10:28:22] Govind Songara left [10:28:23] Alessandra Forti bye [10:28:23] Wahid Bhimji Thanks bye [10:28:24] Queen Mary, U London London, U.K. left [10:28:26] Sam Skipsey left [10:28:27] Alessandra Forti left [10:28:27] Matthew Doidge left [10:28:32] Ewan Mac Mahon left [10:28:34] Stephen Jones left