Software meeting at Mainz on 5 April 2005

Present: Bruce, DaveC, Dimitri, Eric, Florian, Gilles, Kambiz, Murrough, Norman, Pavel, Rainer, Ralf, Stefan, Steve.

The agenda of the main meeting also shows the outline agenda for the software meeting. In practice though we only covered monitoring and calibration with multistep runs. These minutes, written some time after the meeting, contain only some very brief notes from the discussion.

Steve: Monitoring

We had no time in the earlier software workshop for Steve to give his talk on monitoring, so it was carried over to this meeting.

The main points arising from the discussion were as follows:

• the new monitoring working group (MWG) is presently just a TDAQ group but detector representatives should be more involved
• initially this will be done via sending them a questionnaire
• the MWG mandate is for both event based and non event based monitoring (ie the tools should support things like rate meter and histograms from the PPMs)
• the new monitoring tools, GNAM and the PMP presenter are designed for (and by) physicists - but may need some work to improve scalability
• we need to check that GNAM provides support for multistep runs
• there is interest in some common packages (between online and offline) for detectors, eg for bytestream decoding, in addition to the new TDAQ "common" packages
• Q: should we maintain Adrians monitoring software? Depends on the timescale for the new stuff and the effort involved
• Q: can we converge with offline for the 9U ROD bytestream decoding? Probably not yet. Meanwhile we will deprecate all 6U ROD formats except the neutral format.

Calibration

This long discussion was a follow up from that at the software workshop. During this meeting we also referred to some of Florians slides from the main meeting agenda.

The few key points that I managed to note down were:

• calibration runs: we will need to deal with single step runs when running jointly with the calorimeters. These may (LAr) have no markers to distinguish the different steps, though the TileCal do have a method for flagging them
• towards multistep runs: so far we are still using standalone programs. It was claimed we could have multistep runs in two weeks if we did nothing else... we will try
• the main requirements are changes to the module services API and provision of incrementing values at each step via the database
• for the latter we will start with TTCrx settings
• PPM readout via Glink is also a priority though Heidelberg will also need to use readout via VME for local development
• Norman will arrange another joint meeting with the calorimeter groups sometime in May

AOB

• Murrough will have discussions with Richard Hawkings about moving some of our database from OKS to COOL
• Murrough will investigate whether we can use some existing software to control the run control and IS to provide a sequence of runs of different run types.
• We should note that plans for the system management at point 1 are being developed. They may not allow any AFS access or even compilation!