From - Fri Apr  5 13:42:25 2002
Return-Path: <stefano.belforte@ts.infn.it>
Received: from smtp2.libero.it (smtp2.libero.it [193.70.192.52])
	by castor.ts.infn.it (8.11.6/8.11.6) with ESMTP id g34Iu6Y24105
	for <stefano.belforte@ts.infn.it>; Thu, 4 Apr 2002 20:56:06 +0200 (MET DST)
Received: from ts.infn.it (151.25.61.76) by smtp2.libero.it (6.5.015) (authenticated as stefano.belforte@iol.it)
        id 3C99A71000A098D8; Thu, 4 Apr 2002 20:50:17 +0200
Message-ID: <3CACA057.B9CE81AA@ts.infn.it>
Date: Thu, 04 Apr 2002 20:49:59 +0200
From: Stefano Belforte <stefano.belforte@ts.infn.it>
X-Mailer: Mozilla 4.78 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Robert Harris <rharris@fnal.gov>
CC: ALANSILL@tthep2.phys.ttu.edu, r.stdenis@physics.gla.ac.uk,
       badgett@fnal.gov, cranshaw@fnal.gov, Franco Bedeschi <bed@fnal.gov>,
       Al Goshaw <goshaw@fnal.gov>, Robert Harris <rharris@cdfsga.fnal.gov>,
       Frank Wuerthwein <fkw@fkw.lns.mit.edu>,
       Kevin McFarland <ksmcf@fnal.gov>,
       "Todd Huffman (CDF/ATLAS)" <t.huffman1@physics.ox.ac.uk>,
       Jeff Tseng <jtseng@fnal.gov>, bellettini giorgio <giorgiob@fnal.gov>,
       Terry Watts <watts@physics.rutgers.edu>
Subject: Re: Picking up on database export and offsite needs / grid use
References: <020403212605.20a0082d@tthep2.phys.ttu.edu>
	 <3CAC666B.BE18F54F@ts.infn.it> <3CAC7664.A208C5BE@fnal.gov>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thanks Robert,
I understand, and will do as you suggest.
Still, since you are also DH manager, I think it is up to you
and Jeff to say wether for the scenario of largish
computer farms running CDF code around the world, the one single
file catalog is the good solution (then up to the DB meeting to deal
with replica/distribution), or wether a co-operation of distributed
Data File Catalogs (e.g. each made like the one Terry described)
would be needed, and wether a possible adoption of SAM would
make that scenario more or less easy to get to.
Will DB meetings address how I describe a local retracking of a
secondary data set in the file catalog ?
This goes up to more then DH as well.
Once I have a copy of a few secondary data sets and run
complete analysis here, including retracking e.g., I do
not want anybody start argueing about need
to replicate months of work on overcrowded FNAL computers
because they do not trust my copy of the data to be correct.
Who will certify local data copies and code ? You ? Alan ? SAM ?
The usual aisle lobbying by the usual cast of senescent professors ?

I am looking forward to replicate here >20% of CDF data, may
not be possible to just do it over the net SAM jobs after SAM
job, e.g.
				Stefano