Meeting Minutes						      February 07, 2002
-------------------------------------------------------------------------------
Present: Leslie Groer, Pushpa Bhat, Linda Stutte, Florencia Canelli, Ursula
Bassler, Nirmalya Parua, Tom Diehl, Petros Rapidis, Mike Tuts, Drew Alton,
Stu Fuess, Lisa Chabalina, Bob Kehoe
-------------------------------------------------------------------------------
Pushpa made some remarks on the goals for re-visiting the individual detector
examines in detail, and on issues such as Physics Examine, Express Stream,
GUI, Releases and Run Certification.  There were discussions on the importance
of "Run" as an entity for which data quality is required (Tom Diehl) and more
attention is to be given to begin and end runs.  Petros complained that the
data integrity is not being checked, and this can be and needs to be done even
before data is written out.  This was being done to some extent before and
needs to be revived and updated.

During Lisa's presentation where she talked about low-level monitoring tools,
examine, event display and occupancy display, there were discussions of
whether it is possible to archive monitored data flowing in from the control
systems.  DVDQ/chip current data which is displayed and monitored in the
control room is an example. (In fact, there seems to be no alarms implemented
for this either).  The amount of data to store would be too much.  Some
periodic snap-shots might be helpful - if they are stored in a database.
There were also discussions on whether it is possible to set a flag (a stream
of bits) for each event to indicate certain hardware information/status.  For
data coming from EPICS this will not be possible.  More discussions and
thought are needed on what, if any, of the low level hardware status data
being monitored for SMT, needs to be stored.

Florencia discussed calorimeter examine histograms and xgooey usage and
status.  Currently about 90 histograms are available through cal_examine
(monitoring a total of 55,000 channels).  The feedback on xgooey seems to be
positive.  But, this may also have been because cal group seems to be having
problems with histoscope.  There were, once again, questions such as, how do
examines handle runs.  They seem to be run asynchronously and do not know
about runs.  Pushpa mentioned that Geoff Savage is working on providing Begin
and End Run info and Run number info and will provide C++ code that can be
used in examines (calls can be made to his API).  There were more discussions
on ROOT-GUI or xgooey.  No one else except CAL is using it in the control
room.  Pushpa mentioned that it may be worth looking at just a ROOT browser
rather than a GUI that is integrated with D0 framework. This may make it more
efficient.  There will be a meeting on GUI next week.

Currently t01.67 release is being used while t01.73 is available on the online
machines.  The release issue will be addressed by Pushpa soon.

Ursula talked about Calorimeter Data Quality Monitoring.  They have a
Calormeter runs database into which info from monitoring goes.  They also
have plans for offline monitoring using physics/ID groups efforts.  Ursula is
also proposing that CAL shifters look at the histograms produced by offline
monitoring.

Linda showed summary histograms (16) from muo_examine.  These usually go into
D0 e-log.  Muo_examine only processes muon triggers using l3offline_util.
Linda thinks a more controlled event filtering is necessary.  Pushpa will
contact Andrei Mayorov to work on implementing the selection of triggers.  It
was also mentioned that we need more "action" items in examine such as reset,
reset xxx, print, print summary, etc.
-------------------------------------------------------------------------------
Please also see talk slides from the meeting.