Present: A.Cass, M.Cattaneo, C.Gaspar, J.Harvey, B.Jost, P.Mato, F.Ranjard, I.Videau
Minutes: J.Harvey
The minutes of the previous meeting (html) were agreed.
A cooperand was interviewed just before Easter with a view towards working on the Controls project. Currently waiting confirmation he will join us.
LHCb requested three students at the recent TSC round. The LHCb candidate (doctoral student) for the Muon group was selected. Our candidate to work on data management software, Sebastien Ponce (doctoral student), was given free choice between us and a project in LHC division. Lesson : In future must ensure that candidates state clearly their intention to work with us on their application form.
The next meeting of Fellows committee is June 27th. We have at least two candidates who have applied i.e. requested letters of support.
See memo from Beat and Clara.
Beat outlined the programme. One day is reserved for DAQ and Controls issues. Topics were selected that are of relevance to the various subdetector groups.
Financial model : ball park figure for total cost is 650 k CHF +- 100 kCHF. Yearly maintenance fee has two components. The first is the normal maintenance fee which gives right to upgrades, documentation etc and this is costed at about 15% of the discounted total cost i.e. 100 kCHF per year. The second component is for on call direct line support, expected to be ~150 kCHF per year. It has been agreed that sharing of this total cost should be shared between CERN IT / ATLAS / CMS / ALICE / LHCb in the ratio 15/30/30/15/10. Thus the cost to LHCb would be ~100k (purchase) and 15 k / year maintenance. The support to the company would be via IT/CO, who would pay the on-call maintenance cost (~150 k/year). The proposed schedule for license purchase is 30% (2000), 40%(2001) and 30%(2002).
A JCOP agreement on purchase of SCADA system has been prepared and is to be signed by all four experiments. Agreement was reached to launch the call for tender procedure shortly.
The preferred choice is June 27th, with June 28th as a fall-back date. Suggestions for external referees included David Quarrie (LBL), John Deacon and Paul Kunz.
Computing Model : exploitation of Lyon facility,
participation in EU grid
proposal (WP8 - participation from UK, France, Italy, CERN, 1 FTE each over 3 years)
SICb : code management and release procedures to be clarified
and updated on web (Florence), new recovery procedure for faulty cpus(Eric), GOTO
500 events / job (Eric), start major production of bb events with v233 (use Lyon), Profile
and understand performance of SICb (Eric),
BRUNEL - v1 by middle May - Marco, establish discussion forum
and start brainstorming on design of structure/ sequencing of algorithms etc - Marco
Detector Description - screen XML - ?,
'Option A' for managing CDF and XML representations- ?, need simplified geometry for
tracking through material but generated automatically from detailed geometry - ?
GAUDI - doc on web status and responsible of each SICb bank - ?,
improvements to histogram service (overflows, ROOT functionality, fill using ID, forum for
physicists input) - ?, which graphical toolkit (GDK, Qt,...) - ?
GEANT4 - training course (done) - Florence, LHCb projects
(calorimeter,...) - ?
All talks and notes produced by LHCb in response to questions available from LHCb web (html).
Software panel - two rounds of questions answered in two
technical notes. Statements made on behalf of LHCb without chance so far to
consult/discuss within collaboration. Kasemann and White now working on recommendations.
WWA panel - proposal for baseline model documented in technical
note and presented to LHCb by JH during last software week. We have been asked
to provide more precise details .... on what data should be exported and how
(network,...), on backup of data, on MDC strategy, on access to conditions data during
analysis, on cpu estimates. What would be influence on our computing model if 10 Gbps
links became available at affordable cost between CERN and Tier1.
Management and Resources - Computing requirements re-estimated,
scenario for first years datataking with emphasis on calibrating detectors. We are now
asked to provide more information. Software issues : Need a policy statement on
software agreements in view of MoU. Need a model for solving missing manpower especially
for core software. Need to describe first data analysis strategy with emphasis on
reconstruction and event filter. Hardware issues: Strategy for satisfying data handling
and processing requirements i.e. access to resources in Tier 0s and Tier 1's. Emphasis now
is on providing an Interim MoU in which we explain how we could realise a realistic
prototype of computing model in 2003. This should describe the Tier0 and Tier 1 centres
participating, the goals of the prototype (data handling for reconstruction and analysis),
share of responsibilities and costs inside the collaboration, the timescales. It is
also assumed that the final MoU for Computing should be produced in 2003.
A meeting has been arranged for next Monday for discussing these issues in the collaboration.
Feedback from the Computing Review on technical issues should be presented to the TB. Political issues should be presented to the CB.
The next meeting will be on Thursday June 8th.