ATLAS TDAQ Architecture WG. Notes of the meeting of 07-08-02

											GMO 07.08.02





 Present: Bob Dobinson, Saul Gonzalez, Serguei Kolos, Giuseppe Mornacchi


  - Discussion with Chris, Livio and Nick

	The management team was pleased and agreed on the contents of the wg definition document

	The discussion clarified the last step in the "decision procedure":

		* the working group identifies an action to be undertaken
		* the issue and the action is presented to the "management team" ahead
		  of the open meeting where this will be discussed
		* issue & action get discussed, possibly modified and eventually agreed in
		  an open meeting (where the mgt team will be present)
		* immediately following the open meeting the management team:

			1) will formalise the action (who is involved, time scale) if the
			   issue is not a major one, or
			2) will call a meeting of the relevant members of the TDSG to
			   agree and formalise the action

		* the keyword is "timely intervention by the mgt team"

	The mgt team was very supportive of both a "first phase review" and an "architecture workshop".
	It was suggested to:
		1) have a wg status report during the ATLAS week (early October)
		2) have a review (implying the availability of written material) during the
		   TDAQ week later in the year
		3) have an architecture workshop just ahead of the TDAQ review for the TDR

	The formalisation of the working group will proceed as follows:
		- the "definition document" will be submitted to the TDSG
		- the TDSG will officially formalise the wg at the end of september, may
		  be incuding the suggestion of additional members.
		- In the meantime we can start the work.


 - Reading lists

	people are reminded to suggest "intersting" documents for "their" sub-system. Contributions
	received up to now: Online Sw and HLT.

 - Description of the current architecture

	- It is proposed to start the discussions with the various people as soon as possible. If
	  possible, but it seems hard to achieve, during the week of August 26th.

	- A list of topics, which also identifies a preliminary list of components, with names of
	  people to contact (to ask for persons to talk to)

		Component			Ask

		 Level-1/TTC			Nick
		 ROD/ROD crate/ROL		R. McLaren, Ralf Spiwoks
		 ROB/ROS			David Francis
		 L2SV				Bob Blair
		 RoIB				BobBlair
		 Networks			Hans Peter Beck
		 DFM				Hans Peter BEck
		 SFI				Hans Peter Beck
		 EF sub-farm			Chris Bee
		 L2PU				Chris Bee
		 SFO				Hans Peter Beck
		 Online				Mihai Caprini

	 Apart from online the components above correspond to hardware and software.
	 As for online it was agreed that:
			
			1) the architecture of the online software is not in the scope of this wg
			2) we should discuss the functionality, the integration and coordination roles
			   of the online software
			3) Eventually the working group should work in the direction of identifying
			   missing functions and/or identifying where rationalisation for the use
			   of online sofwtare can be introduced.

	Each item in the list should take an average 2 hour slot.

	We should provide the people with some guidelines of what we are interested in discussing. For example
	burning issues, issues on the critical path, and may be questions extracted from our work list. It
	was agreed to circulate by e-mail, before theend of this week, suggestions on "topics to discuss" which
	could form a preliminary "guideline".

	- Sharing of contacts:

		Beniamino for LVL1/TTC/ROD/ROL
		Benedetto for ROB/ROS
		Bob	  for data collection
		Saul	  for L2Pu and EF
		Serguei   for online

         will get in touch with the contacts listed above to find who we could talk to and when. May be a draft
	 calendar of discussions could be set up during the next two weeks. Taking into account that:

		giuseppe is away 12-25/8
		Bob is away the 1st week of september
		Saul is away the second week of september
		others.. ????

 - Parameters, standard values

	Beniamino has started collecting information related to the detectors

 - Database of information

	The possibility of creating a database, in an as yet unknown form, of the information collected by the
	working group is being explored. The aim is 1) to collect, for each component, the relevant requirements,
	pointers to documents, parameters and their values 2) to have an interactive (web based) drawing of the
	architecture to explore/edit/display the contents of teh data base. The same data base could also be
	exploited to produce a costing model, as input to paper models, to produce lists of parameters, requirements
	etc.
	Some small prototype will be done to understand the usefulness of such a tool

 - Use cases

	* They should not be seriously tackled before the tour of the current system is completed.
	* giuseppe will distributed some reference material on use cases (what should be in etc.). To see how
	  possible it is to formalise the contents of use cases.
	* already when discussing with people we could suggest that they think about use cases related
	  to faults and tests
	* How should we address use cases? One way could be to suggest the creation of ad hoc "task forces"
	  with a suitable composition, including members of the wg, to analyse separate use cases. 

	* Histogramming scenarios will be added to the list of use cases

 - Milestones etc.

	* we should start to put some milestones (reviews etc.) in the table at the end of the document:

	
 - Next Meeting: it is suggested to have th enext meeting on Wednesday Aug. 28 at 16:00