Attendees ========= Derek Ross Pete Gronbech Sam Skipsey Richard Hellier David Crooks Daniela Bauer Mark Mitchell Brian Davies Evan MacMahon John Bland Andrew McNab Stephen Jones Stuart Purdie WInnie Lacesso Graeme Stewart Duncan Rand Raja Nandakumar Christopher Walker Govind Songara Santanu Das Experiments problems/issues ============================ LHCb ---- MC production started UCL banned - not known why LHCb model may be changing next year - more reconstruction going to Tier 2s? Probably select Tier 2s (6-10), files uploaded to Tier 1 so no disk increase at Tier 2 Atlas ----- RAL Problem - lot of failed transfers with Castor due to load issues due to disk server draining Draining has been reduced from 3 servers to 1 per service class Doing very well in MC production but running out of space at Tier 1 - will migrate some files to tape CERN Power cut Cambridge power cut problem fixed Durham have network issues Try Site Status Board Thanks to all for very good year in Atlas Computing CMS --- No CMS representative Moving to CREAM CEs in Jan Other ----- Nothing Meeting and Updates =================== ROD update ---------- No issues on dashboard when DB took over GridPP Nagios not available? Intermittent Network issue? Tier 1 Update ------------- Quiet Security -------- Daily/Weekly CSIRT e-mail is only sent to CSIRT team list, Jeremy and Alessandra on list Once pakiti detect vulnerability on site send e-mail to list, and security officer on duty will open ticket on site (copying Mingchao) Pakiti link https://pakiti.egi.eu Focusing attention on critical and high risk vulnerabilities. Other vulnerabilitis left to site admins Need security officer role to view pakiti page Site admins should contact Mingchao for access to pakiti page if they can't access it Developing a dashboard for security issues, available at some point next year GDB --- CERN taking ownership of orphaned middleware required for WLCG Increasing popularity of cvmfs, QMUL and Tier1 leading. IPv6, Security group VOs talks interesting EGI user forum WLCG workshop Site Updates ============ Glasgow ------- not doing anything critical, 1/2 new disks in. Accepted new hardware CREAM - 3, no issues Imperial --------- Power Cut - lost 1 disk server Have CREAM and APEL for ages Bristol ------- Have cream CE, still have to do glite-APEL Oxford ------ Problems on 1 disk server, gridftp and rfio process died, were restarted with no issues. New kit is in. WN are working, Disk servers are in testing CREAM CE - 1 (+ 1 lcg-CE) will make another CREAM CE in the future glite-APEL - in place Liverpool --------- Things going okay, power outage caused some days outage 32bit nodes scrapped Cream CE - Yes glite-APEL -Yes Manchester ---------- CREAM CE - working okay apart from test fails at end of week but starts working over weekend, possibly due to farm usage Test jobs are priortised over others, so not the issue glite-APEL - yes Brunel ------ Cream CE working on APEL Fixed CMS DPM issue UCL-HEP ------- Power maintenance over w/e, just switched back on, having nfs issues - skeletal cover Installing CREAM CE- unsure of status APEL - probably not QMUL ---- Disk server order placed, delivered before new year CREAM CE - up and running APEL - being done today FTS issues on Monday, channels increased to RAL Power outage on 18th Jan - coincides with panda downtime - to enable voltage optimiser RHUL ---- Site running smoothly - no issues glite-APEL in production, in touch with developers, have not been publishing for last 2 weeks CREAM CE installed but not yet configured or tested Lancaster --------- glite-APEL - installing CREAM CE passing SAM tests, will be doing proper wokr for Atlas once software installed ECDF ---- All running smoothly CREAM and glite-APEL in advanced state, will be done early next year Birmingham ---------- Has cream and apel Cambridge --------- Powercut last two days CREAM CE to be done glite-APEL to be done tomorrow RALPP ----- CREAM CE glite-APEL to be done this week Lot of new kit, some running, some doing acceptance tests AOB === DTEAM Voms server configuration - VOMS_SERVER line for dteam in site-info.def must not contain CERN, but should remain in VOMSES line GGUS ticket to update once changes have been made, https://gus.fzk.de/ws/ticket_info.php?ticket=65323 Chat Window =========== [10:50:58] Richard Hellier joined [10:50:58] Richard Hellier left [10:54:13] David Crooks joined [10:56:24] Daniela Bauer joined [10:57:51] Mark Mitchell joined [10:59:27] Brian Davies joined [10:59:59] Ewan Mac Mahon joined [11:00:07] John Bland joined [11:01:41] Andrew McNab joined [11:01:52] Stephen Jones joined [11:02:26] Stuart Purdie joined [11:02:39] Winnie Lacesso joined [11:02:47] Graeme Stewart joined [11:03:01] Duncan Rand joined [11:04:03] Raja Nandakumar joined [11:05:15] Daniela Bauer Dave Colling sends his apologies. He's got another meeting. Stuart has not yet resurfaced after last night's Christmas party. [11:08:02] Christopher Walker joined [11:09:49] Govind Songara joined [11:10:37] Sam Skipsey Obviously you lot party harder than we do in Glasgow, Daniela. We were boringly responsible at ours. [11:13:15] Sam Skipsey (Of course, the other solution is to use only ARC "CE"s... ) [11:13:38] Santanu Das joined [11:21:01] Christopher Walker CVMFS: It now syslogs when it has failed to download a file it wanted. [11:21:43] Christopher Walker CVMFS: If LHCb or CMS want it enabled, QMUL is happy to turn it on here. [11:23:50] Mingchao Ma joined [11:24:13] Mingchao Ma sorry, was in EGI OMB meeting [11:24:29] Chris Curtis joined [11:26:04] Winnie Lacesso . cream-ce is ok, still have to do glite-APEL. [11:27:32] Chris Curtis left [11:29:23] Ewan Mac Mahon You did well to avoid having to pay someone to take them away. [11:30:46] Christopher Walker Who paid you? We've still got a few kicking about... [11:31:08] John Bland Concept management limited [11:31:22] John Bland www.conceptmanagementuk.com [11:32:02] John Bland they do weee recycling but will sometimes share the profits to secure the deal [11:40:03] Ewan Mac Mahon Just had to pop of for a mo there; did I miss anything? [11:45:34] Pete Gronbech https://pakiti.egi.eu [11:46:02] Raja Nandakumar left [11:46:29] Pete Gronbech sys admins need to apply for the site security role on the gocdb to be able to see the pakiti page [11:46:49] Pete Gronbech Best to also email Jeremy to ask him to approve the role. [11:47:07] Ewan Mac Mahon Raja's not in the same building as any computers, is he?..... [11:47:29] Derek Ross only some of the Tier 2 [11:47:38] Derek Ross [11:51:17] Pete Gronbech https://gus.fzk.de/ws/ticket_info.php?ticket=65323 [11:51:30] Sam Skipsey left [11:51:30] David Crooks left [11:51:32] Winnie Lacesso left [11:51:32] Richard Hellier left [11:51:32] Govind Songara left [11:51:33] Graeme Stewart left [11:51:34] Mark Mitchell left [11:51:35] Brian Davies left [11:51:35] John Bland left [11:51:36] Winnie Lacesso joined [11:51:36] Andrew McNab left [11:51:39] Duncan Rand left [11:51:41] Ewan Mac Mahon left [11:51:43] Mingchao Ma left [11:51:49] Stephen Jones left