Presentations:

Harvey Newman:       "Regional Centre Committee"
Ian Willers:                 Review of RC Requirements
Ken Bell:                   "RAL as CMS Regional Centre?"
Luciano Barone         "Concept and Roles of RC"
 

Harvey Newman:
Harvey raised questions in his talk about what people's views of the architecture of Regional Computing Centres are.  What kind of disk and network resources are needed?  What software?
Should the different centres be forced to be architecturally similar to each other?
What prototyping is needed?  What are the roles of centres at institutes with smaller CMS commitments, such as Pakistan or Rutherford-Appleton Labs?

In the discussion, the point was raised that, in order to answer these questions, we have
to first examine the tasks of these centres, such as analysis, simulation, production, and
software development.  It was also debated about whether it was premature to request
money for prototyping at this stage, even if it comes from institute budgets, not CMS funds.
 
 

Ian Willers:
Ian summarized the "current thinking" about regional computing centres.  The computing
power at each centre should be at least 10% of that available to CMS at CERN, and
the funding for the centres should come entirely from the institute itself.
He listed some candidate centres around the world.
 
 

Ken Bell:
Since the UK has three ATLAS physicists for every CMS physicist, it doesn't seem likely
that the UK's CMS contigent can sway enough funding to make a CMS Regional Centre
with the aforementioned 10%-of-CMS-computing-power at Rutherford-Appleton Labs.
Maybe they could manage 7%.  But it looks like RAL can be an ATLAS regional centre.
So the question is, would it be useful for us to make a little "outpost" centre in the UK?
 

Luciano Barone:
The INFN model of a regional centre is a "virtual" one, distributed across all the
institutes in INFN.  The "cells" look very similar to each other, and access to the cells should be
transparent (a la AFS), or, ideally, through something like a Web interface.

In his discussion of the roles of a regional centre, he emphasized the need for bookkeeping.
 
 

Discussion: Letter of Intent to LCB?
Next came a discussion about whether it made sense now to write a
Letter of Intent to the LCB to investigate the problems related to
distributed computing.  The INFN groups were interested in at least
defining the project, so they will get in touch with Kris Sliwa from ATLAS.
 

Subgroups for the various RC types:
The following people agreed to lead investigations into the roles of Regional Centres:

Production:  Luciano Barone
Analysis:      Irwin Gaines
Simulation:  Hafeez Hoorani
Software:     suggested was Vincenzo Innocente