NUC 13.10.2010

Attendance:

Hartmut Stadie, Andreas Gellrich, Andreas Haupt, Steve Aplin, Yves Kemp, Wolfgang Ehrenfeld, Marcello Barisonzi, Johan Blouw

News

December meeting of HGF, contributions?
Grid centre review: presentation per experiment
face-to-face meeting: January 12, 2 pm to 6 pm

NAF status

AFS client patch partly deployed
- partially successful
- some problems remain, but might be simple overload
- keep reporting problems with AFS
update on action items (see below)

Existing Action Items

id	who	description	status/comments
	old-1	NAF/ATLAS	CMT problem	looking into local software installation
	1005-1	NAF	present SL4 work group server usage	no SL4 node remains closed
	1005-3	NAF	email notification for /scratch monitoring (instance full)	should be ready end of next week
	1006-6	NAF	dCache upgrade time line (10 GE in HH)	80 % of storage connected, waiting for new hardware at end of year
	1007-1	NAF	setup Lustre groups for ATLAS, CMS, ILC	problem can be reproduced, have to report bug, takes time
	1007-2	NAF	report on help desk tickets	shown in NAF status report
	1007-3	ATLAS/CMS/LHCb	ICHEP review; get feedback
	1010-1	NUC	face-to-face meeting	in January, afternoon
	1010-2	NUC	resource requests
	1010-3	NAF	automatic warning about AFS quota	first version of scripts exist, warning at login
	1010-4	NAF	statistics on multicore/PROOF usage	proof: 570 wall days; mulicore: 24 WALL days in last month, wewant to have this regularly
	1010-5	NAF	update News section on webpage	closed

ATLAS report

Accounts: New accounts (4x)
ATLAS Software: Requested more AFS space for software and DB release installation
Transient problems:
- common AFS problems
- AFS token problems
- dCache problems
  - overload with one or two pools (LOCALGROUPDISK) everybody affected because of file distribution strategy
- dCache downtime:
  - problems with conditions access (files on Lustre)

CMS report

people unhappy with storage reliability
dCache user dirs need better quality, maybe replication -> new action item
Lustre hangs? -> report problems

ILC/CALICE report

no problems

LHCb report

AFS problem found for SGE job

New Action Items:

1011-1

CMS

strategies for dCache user directories

Next meeting

- Wednesday, December 8th 2010, 1 am