Software phone meeting on 11 May 2006

Present: Bruce, Damien, Eric, Florian, Gilles, Kambiz, Murrough, Norman, Paolo, Pavel, Rainer, Richard, Victor, Weiming.

Status reports

Birmingham

Gilles has been configuring his system to include the 9U ROD crate from RAL. He has been having problems preparing the boot image for the CPU (a different model from the others in Birmingham) but is nearly there after getting help from Bruce.

Bruce remarked that we should upgrade the boot EPROMs on the older CPUs to allow a common boot protocol, but has had no time to do this.

Heidelberg

Florian has finished his run plan handling code and is now writing documentation. He now wants to prepare the setup for doing a real cable test. Victors monitoring should be changed to take events from the monitoring stream not buffer files so that it is active during the run. Florian will interact with Victor.

Florian also reported that there were problems with CVS access this morning. But by the end of the meeting these appear to have been fixed.

Victor has finished his PPM monitoring document. He will discuss with Murrough where to store it.

Mainz

Rainer said that the first preproduction JEMs should arrive in Mainz next week for tests.

A meeting on ROD-ROS mapping was announced recently. It was not clear if we should be involved given the recent interest in PPM data at level 2 for jet/energy triggers. However Rainer said that studies suggested the PPM data would not be that useful so we should not be too bothered. Norman will check what the scope of the meeting will be.

QMUL

Paolo has been adding more histograms to gnamL1CaloHisto and will commit the new features soon for people to try.

Murrough reported that he had made most of the necessary changes to our packages to build under the next TDAQ release and gcc344. He has also restarted looking at accessing the COOL database from run control using the new CORAL based version of COOL.

RAL

Norman and Richard have been reconfiguring the RAL system without the 9U ROD crate that is now in Birmingham.

They have also been chasing the reported problem of CMM registers losing their contents. Tests where registers were continually read by VME clearly showed this behaviour. Ian is now looking at the firmware. Norman was worried there may be the same problem with other registers and modules. In particular they will try exercising the TTCrx I2C connection.

Damien, the new colleague at RAL, is going to start looking at the effectiveness of the few favoured PPM compression algorithms with realistics Physics data from Alan.

Richard has continued looking at tools to "poke and prod" the database. With the help of the author (Federico Zema) he is now running COOL_IO outside CERN. However progress with KTIDBExplorer is not so good and he has received less help from its developers.

Stockholm

No report.

CERN

Bruce has made the long pending updates for formatted buffer readout. Documentation is to follow (maybe after reminders!).

Kambiz described the changes to PPM firmware he is making to improve the behaviour of the VME readout in multistep runs, especially when running with the LAr at trigger rates above 30Hz. This involves enabling/disabling readout into the FIFOs and also providing a fast clear for the FIFOs to avoid having to completely read them to empty them.

Database Workshop

In preparation for the workshop (22-23 May) Norman reviewed the list of database areas - he will send round a list with names of people who should try and be knowledgeable in each area.

Points from Run Structure Workshop

Murrough mentioned the recent run structure workshop. In particular Benedettos talk on "Stopless recovery" might have an impact on us if we need to enable/disable modules or their inputs during a run.

ROD formats for PPM

Norman said the subject of Slink ROD formats had been raised (again) in particular the handling of numbers of slices and restrictions on these imposed by the ROD. He had a proposal to adjust bit fields in our Slink subheader word to address concerns and allow larger numbers of especially FADC data to be read from the PPM without affecting the parts of the subheader field used by other modules. However no decision was taken.

Contacts during TDAQ week and after

Norman will arrange meetings with the RoIB people and also the calorimeters on calibration issues. There also needs to be a private discussion on ROD formats since the above was not conclusive.

Next meetings

The next phone meeting will be on Thursday 25 May at 11:00 CET (10:00 GMT).