PrimEx Software Meeting Minutes
Thursday, April 8, 2004
Present: Eric Clinton, Dan Dale, Raphael Demirchyan, Mark Ito,
David Lawrence, Dustin McNulty, Vasily Mochalov, Itaru Nakagawa, Aram
Teymurazyan
Nominate secretary
Mark agreed to act as secretary.
Approve last week's minutes
Approved as written.
Eric reported that the problem with the "double" beam hole that
Dustin showed at the last meeting was traced to an indexing problem in
psim_digitize and has been fixed. He mentioned that there are other
issues with clustering that still need to be resolved.
Changes to today's agenda
None.
Readyness and schedule for online test II: Itaru
Itaru announced that we are tentatively scheduled to begin the second
online software test on Monday, April 12 starting at noon. Sergei has
not made a firm commitment to giving us this time, so Itaru encouraged
us to be flexible.
Dave mentioned that if anyone would like to test their code in
advance of Monday's collective test, the DAQ system in the Test Lab is
available. It has all of the necessary components, including a
functioning ET ring, that are necessary to run our online software.
Itaru asked about the readyness of the veto monitoring
program. Eric reported that veto_calib is close to finished, but there
might be a few bugs left to squash before Monday.
Itaru also asked about readyness of the online HyCal calibration
code. Raphael reported that there was still some work remaining in
the conversion to ROOT.
Dustin mentioned that the ability to look at merged events (one
event from two "minimum bias" events) to simulate accidentals in the
data stream would be an significant advance over what we did in the
first online test. Dave agreed to point Mark in the right direction
for fixing the code to do this.
Dustin mentioned that the pi-zero events that we ran with last
time had the bug (mentioned above) in psim_digitize reflected in the
data. We could re-digitize this data in preparation for Monday. Itaru
and Dustin will make a decision on whether this is worth the effort.
Itaru agreed to write down an informal run plan for the test on
Monday and send it to the email list.
Temperature Monitoring for HyCal: Jar
Jar showed us a live display of his temperature monitoring screen. Here is a screen shot from yesterday:
The scheme is to use a web browser to access the PC's console display
via VNC. The PC itself will be located in the Hall. Jar is working
with the computer center to install a remote reboot facility.
The group thought that what he had was a pretty good
representation of the quantities that we should be monitoring
visually. Still there were several suggestions:
- For the items like input temperature and output temperature,
label the temperature indicator with the appropriate text.
- Label the x-axis on the strip charts with time rather than sample number.
- For the strip charts show a wider range of temeratures so that
sudden changes will not be auto-scaled away (if auto-scaling enabled)
or disappear off-scale (if auto-scaling is disabled).
These data will be available as EPICS variables as well. This will be
necessary to insert them into the datastream. In addition, we should
set up EPICS-based alarms. It will not be necessary to re-create the
visual display in EPICS.
Status of non-expert online histogram-viewing system: Aram
Aram proposed a system for viewing our online histograms. The ROOT
macro he has written will have two modes, manual and auto, manual for
experts who want to select individual historgram, auto for the
non-expert who want to view a pre-selected subset (not quite what he
presented, but close). The user configures the system by specifying
the systems that should be available for display. The configuration
can be saved and will be used as the default for subsequent sessions.
Some suggestions to Aram:
- Dave was worried that the technique of fetching all histograms for
selected systems might prove prohibitively slow.
- Vasily thought there there should be a facility for the expert to
navigate the various histograms that are available in order to track
down anomalies in the data.
- Dustin thought that the gui should present system choices to the user
via a pull-down menu.
- Dustin suggested that the system should be able to process
pre-written ROOT macros as well as display individual histograms.
- Itaru cautioned Aram that he needs to be careful with ROOT
software versions.
Notwithstanding these and other excellent suggestions, initially Aram
will focus on a minimal system for the non-expert shift taker.
Standard root-tree/ntuple for PrimEx: all
Mark brought up the idea of having a standard PrimEx root-tree or
ntuple that we could all use. That might avoid duplication
effort. Dave thought that this was an issue that we could defer until
we have some real production data to work with. We went along with
Dave's suggestion.
Beam normalization software issues: all
Mark wanted to explore whether we need to have some high-level
planning to insure that all of the necessary software components to do
flux normalization are in place. If a plan is called for, how do we
develop the plan?
Dave reminded us that he and Itaru have been working on software to
do online photon flux monitoring. The work has been on hold recently.
We deferred a full discussion of this item to a future meeting given
the lateness of the hour.
Software meeting time: all
We agreed to move the meeting to Tuesdays at 3:30 pm starting the week
after next. We felt that having a software meeting this coming Tuesday
would be too soon after the online test.