Minutes of Online Calibration Meeting 10/27/00

MINUTES OF ONLINE CALIBRATION MEETING 10/27/2000

Theme: 'Are We There Yet?'

   - Have we accomplished our calibration goals for the
       commissioning run?
   - What's left?
   - Can it be done next week?
   - Are all the calibrations integrated into the shift work?
        Any feedback?
   - What is the status of documentation?

Calorimeter (P. Gris/J. Cranshaw):

Philippe reported that the LED/XEF calibrations are part of
regular shift work, but he needs some new tables to make the
LED/XEF calibrations fully functional. He needs a table with
the D-mode pedestals for the CEM with no flashing, and he
needs a table with the pedestals from the readout of the PIN
diodes. This will finish the work of getting the systems fully
tested and working and can be done next week. He needs the
help of B. Badgett and J. Cranshaw for downloading and
database aspects respectively. Philippe maintains a LED/XEF
user's guide on the ace page.

The Qie calibrations have no outstanding issues, although the
version of the code running online needs to be upgraded to
correct a core dump problem and to properly include the CLC.

PHILIPPE COMMENTED THAT THE CO'S HAVE NOT
BEEN PROVIDING MUCH FEEDBACK BECAUSE
THEY'RE NOT SURE HOW TO IDENTIFY PROBLEMS.
ONE OF THE KEY FEATURES WHICH MUST BE ADDED
TO THE ONLINE CALIBRATION SYSTEM IS COMPARISON
OF DATA/HISTOGRAMS TO REFERENCE
DATA/HISTOGRAMS.

CLC (absent):

Silicon (J. Slaughter):

For silicon, things seem to finally be coming together. The
downloading of pedestals and suppression of dead/hot channels
was successfully tested on 10/26. At this point, they would
like to start writing to the production database rather than
the development database. Also, they will be working on
stabilizing their procedures. They do not use their consumer
online, but rather analyze the data offline and designate
which set to use online. They are also using the Sets part
of the database schema and have established a link between
online silicon calibrations and offline silicon tracking.

D. Box commented that he has been working on implementing
float arrays which would be useful for correcting some design
flaws in one of the silicon tables as well as in all the Qie tables.
Once silicon has moved out of development, we should try
these out in development ASAP. This brought up the question
of table naming for changes to existing data tables.

COT (B. Orejudos/A. Kotwal)

Bill reported that the front end timing calibrations are documented
and part of the shift work. Unfortunately, they are not producing
reliable numbers due to problems with triggering which require
access and modifications to the TRACERs. It was not foreseen
that this would be done during the commissioning run, since they
are trying minimize changes to the detector at this point. The
connection to tracking is being tested, but it currently bypasses
the database totally. Bill and P. Tamburello were going to work
that out. In addition to timing, dead/hot channel suppression
is being used in the tracking. There seemed to be a bit of
confusion on whether the front end timing calibration was
to be used at Level3. Also, it was unclear what part of Stage0
was used by Level3, so Bill volunteered to talk to K. Burkett,
and drag him to the next meeting.

Ashutosh was unable to connect by phone, but he commented
by mail. His comments essentially paralleled Bill's comments
on the status of the timing calibrations.

SMX (absent):

Software (J. Cranshaw):
Jack reported that the consumer template GenericDMode3
has been upgraded to send messages to the ErrorLogger.
Also, some of the the consumers are still using GenericDMode2,
so they should be upgraded before the end of the commissioning
run.

Operations (R. St. Denis):
Various problems had been encountered. The most serious
of which was when we could not switch over from the production
database to the integration database because someone had
hardcoded the production db into their online code. The conclusion
was that more supervision of online database access
was needed.

THERE WAS ALSO A CONSENSUS THAT MAINTAINING
SHIFTS DURING THE WEEK AFTER THE BEAM STOPS
WOULD BE USEFUL TO THE CALIBRATION GROUPS.

NEXT MEETING 11/10: 'Are the Tires Flat?'
  - Review of consumer and DBANA frameworks and functions.