Core Minutes 4/27/2004
Present: Joanne Bogart, Toby Burnett,
David Chamont, Jim Chiang, Seth Digel, Richard Dubois, Dan Flath, Berrie Giebels, Navid
Golpayegani, Traudl Hansl-Kozanecka, Heather Kelly, Michael Kuss, Matt Langston,
Julie McEnery, Pat Nolan, James Peachey, Dirk Petry, Sean Robinson, Bob
Schaefer,
Alex Schlessinger, Tom Stephens,
Tracy Usher
-
Core issues and status: (Toby) See
Toby's
run-down of issues of interest
here;
additional discussion below.
- Reproducibility: (Michael)
He has
documented his extensive investigation. Observations include the
following:
- using GlastRelease v4r1 and simple sources, he was able to reproduce
events as long as the time for the first event is correctly specified (time
matters because of the motion of the spacecraft), however
- running with the gui changes things; it always uses a start time of 0, so
cannot in general reproduce events simulated without the gui (Toby) Should be
able to fix this
- the story is more complicated with composite sources. Here history
comes into play in the choice of sub-source. As a work-around, one can
generate for each subsource, whether or not is it the winning one, then
discard the ones which haven't been chosen.
- using GR v4r2, he found a separate problem with the muon_pencil_angle
source, but not for all_gamma. This appears to be connected with
TkrDigi::addElectronicNoise (and not at all with random sequences)..perhaps a
bug?
- Release Manager:
- Alex expects to have a new version, supporting both optimized and
unoptimized builds of releases, in place in a day or two. He'll also be
moving output to a different disk.
- (Toby) how about producing .tar.gzip of source as part of the build
procedure? Would save remote developers a lot of time. (Alex) OK.
- Navid has written an algorithm to check how much time is remaining,
allowing over-long jobs to exit with some grace..unless the problem is a
single super-long event, since the algorithm is only invoked between events.
- Fix LATEST! Owners
of G4Propagator, AcdRecon,
mcRootData, digiRootData,
reconRootData,
commonRootData please note: LATEST won't build until all of the above
have been modified to use GlastPolicy v6r*
- Install area: (Traudl) She and Heather
working together have put changes needed for the install area into some of
Heather's packages, such as AcdRecon. Serious migration can't proceed until
the new CMT and Gaudi versions are installed. (Alex) will try to do CMT
today.
- GlastRelease v4r2: (Heather) all_gamma
runs are done; backgrounds next up once she is sure we have the space for
them. Is 50 million enough? (Richard) Bill would probably want another
factor of 10, but we surely don't have space for that. (Toby) Plots from
all_gamma data look good.
- Science Tools: Current HEAD is about
ready for release (James) Recommends destructor as shown towards the bottom of
Toby's status page for virtual base classes.
- InstallAnywhere: (Pat) Turns out the free
version he was using doesn't get along with Redhat Enterprise. He had to start
over with one of the two commercial versions; hasn't yet achieved all the
results he had with the free version.
- External libraries for 7.1:
The version Matt
made earlier needs G4 and EBF. He will incorporate them, probably today.
- New technology: (Richard, Matt) On
Monday Matt gave a presentation (pdf,
ppt) and demonstration
concerning J2EE and how we might use it as the platform for several key
applications, including the data processing pipeline and system tests. In
addition to the usual GLAST suspects, Tony Johnson (local expert on all things
Java, though not specifically a J2EE jockey) attended; we hope for continued
consulting help from him. J2EE is an increasingly common technology for
business applications, but new for HEP. Other groups at SLAC, such as
(accelerator) Controls, have also expressed interest in it.
Since this is potentially a Big Deal we should keep track of
how the process is going (maybe a Wiki page?) As a test case, Matt will use it
to build a new interface to the system tests; it should be a fine way to
exercise the advertised legacy-wrapping facilities. He estimates this will
take him a week to 10 days.
(Seth) Why not continue using OPUS for the pipeline? (Dan et. al.) We have
no use for two of the major chunks within OPUS (database, batch) because we
are committed to well-supported local alternatives (Oracle, LSF). This was
known from the start. What Dan subsequently discovered is that the parts of
OPUS we're not using are actually getting in the way: he has to go to extra
trouble to code around them.
-
Flight Integration: The
weekly run-down:
- EBF/LDF news: (Heather) There have been two more tags of LDF. We
have so far upgraded to the first one; will move to the latest shortly.
- 2-tower simulation: (Heather) She has so far been unable to
reproduce the magic combination of package tags Leon used a month or so ago
for his missing tower work.
However, he is due back soon.
- Calibration database interface: (Joanne, Marco, Riccardo)
Riccardo has started work on the gui for the insert function; Marco
continues to work on the query interface.
- Release manager builds for engineering model: Alex says this can
go in with the imminent upgrade.
- AOB - fuzzy clusters: (Tracy)
It builds, it runs! Hence it can now readily be explored to see whether it's
suitable for us.
- Document effect of MC pruning options (Riccardo) (8/5/2003) It's
documented in the code; Tracy will look into putting the information in a
more accessible place. (9/2/2003)
- Write up, and perhaps present at a future meeting, hints on debugging.
(02/18/03) Some combination of Toby and Tracy will provide
something for Windows.(7/22/03)
- Overhaul documentation altogether by about November (Heather and ??) (9/2/2003)
In progress; not yet done (01/13/04)
- Resolve calibration/timestamp issue. (11/20/02) Online folk have not yet thought about
putting timestamp information into EM data, but, thanks to Richard's
inquiry, may now do so. (2/25/03) We can probably live without a realistic
system for EM, but we will put it on the agenda for the DC1
workshop.(5/6/03) At the DC1 Workshop Steve Ritz promised to find an
authoritative reference for timestamp format and pass it on. (7/17/03) Now
that events have "real" timestamps, Joanne needs to reassess
situation with Calibration (9/2/2003) Timestamp exists in EM data. Needs to
be fetched by Calibration (1/13/04)
Put on hold until formats settle down. (3/16/04)
J. Bogart Last Modified:
01-Jun-2010 15:46:20 -0700
previous
