From: Saul Gonzalez [Saul.Gonzalez@cern.ch]
Sent: Tuesday, July 23, 2002 5:07 PM
To: Bob Dobinson
Cc: Giuseppe Mornacchi; serguei.kolos@cern.ch;
beniamino.di.girolamo@cern.ch; Benedetto Gorini; reiner.hauser@cern.ch
Subject: Re: FW: AWG ideas


Hi -

Here are a few thoughts on the possible role of the AWG:

* I think the role of the AWG should be more of a "catalyst" for
  developing a coherent architecture. To this end, it is essential
  to constantly consult with the various sub-systems and to start from
  the present view (as discussed at the last tdaq week).

* As was evident during the review week, one of the weakest areas
  in the TDAQ is in system-to-system interfaces. The AWG should focus 
  on this and make sure the right people are talking to each other.

* Another critical area are external dependencies. For example, in the
  HLT we now have a strong dependency on Gaudi/Athena. A formal mechanism 
  to track this should be in place, since it has serious implications on 
  the HLT's software architecture.

* To the extent that testbeds validate an architecture, who coordinates
  and prioritizes this work? Is this within the scope of the AWG?
  Related to this point, what metrics will be used to validate the arch.?

* The global optimization of TDAQ, in terms of cost/performance, should 
  also be tackled as soon as we have a 'credible' architecture. Here a 
  good L1/L2/EF model will be needed, as well as close interactions with 
  offline/physics communities. It would be nice to be able to show that
  our arch. allows us to re-configure resources (sw+hw) easily according 
  to run needs.

best regards, Saul