Minutes of Online Meeting
27th May, 1998

Agenda:

1. News
2. Progress of Readout Unit
3. Progress on DCS Project and Planning
4. AOB

1) News

• Frank Harris's DAQ convenor mandate ends end of June
• Referees coming in July require more detailed writeups
  -> FE/TTC application writeup B.Jost

2) Presentation on "Eventbuilding Strategy" by J.P Dufey (slides)

Goal: Project to study LHCb eventbuilding

• Effective throughput
• RU network
• SFC eventbuilding

Commercial switches: non-blocking but not large enough
Network bandwidth: Theoretical figure of max. packet switching time for point-point
Networks is 2-SQRT(2) = 60 %. For Myrinet, 30-60 % was measured
General approaches to solve traffic contention:

1. flow control ( FiFo)
2. collision detection

ATM fix is traffic shaping (multiple FiFos) based on two Approaches: barrel shifting or rate division
close to 100% BW usage can be achieved for ATM
Readout Unit: need consider data flow in relation with the network
Data structures for subevents to be defined containing EB control
Frames added by network interface add overhead .
Recommendations needed both for RU and SFC ( URD's )
Simulation of all architectures and technologies: ATM, SCI, Myrinet, Gbit ethernet, other

Some comments:

J.Harvey: evaluate and prototype to build knowledgebase. Gain experience, then design &build
H.Muller: RU project was presented before with timetable to freeze design in 2003, after preselection ( 2001 ) with prototype
T.Nakada: reduce number of network technologies rather than augment
P.Mato: which are the design criteria: maintenance certainly important
B.Jost: EB project scope, look more at EB problem. Links and institute participation open issue

3) DAQ meeting planning for 15th

• RU project ( H.M)
• Output stage RU ( JP.D)
• FEMs (B.J)
• Supervisor
• Link technologies ( check w . M.Letheren)
• Controls (P.M)
• TTC ( B.J)

4) AOB

Next meeting: J.P. Dufey Project proposal

Cooperant Requests (use Aleph and Group channels)

Minutes by Hans Mueller