GlueX Offline Meeting, October 31, 2012
GlueX Offline Software Meeting
Wednesday, October 31, 2012
1:30 pm EDT
JLab: CEBAF Center, F326/327
- 1 Agenda
- 2 Communication Information
- 3 Minutes
- CCDB creating SQLite files nightly
- CentOS6 farm nodes removed from farm this week
- Review of minutes from the last meeting: all
- Report from the last Data Challenge Meeting
- Anomalies in MC units: Elliott
- hddm2root: David
- Resources: David
- Reconstruction sub-group reports
- PID: Reorganized Offline HOWTO List
- Action Item Review
- Review of recent repository activity: all
Talks can be deposited in the directory
/group/halld/www/halldweb/html/talks/2012-4Q on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2012-4Q/ .
- CMU: Paul Mattione, Curtis Meyer
- JLab: Mark Ito (chair), David Lawrence, Dmitry Romanov, Simon Taylor, Elliott Wolin
- There is a cron job creating SQLite versions of the CCDB nightly. Find them at JLab in /group/halld/Software/calib/ccdb_sqlite .
- CentOS6 farm nodes removed from farm this week for a linpack test. The test should be over on Thursday.
Report from the last Data Challenge Meeting
Mark created a web-accessible directory for keeping track of data challenge conditions.
Anomalies in MC units
Elliott reported seeing anomalous units in some fields in the raw data. Energies should be in GeV and time in nanoseconds and he is seeing other values. Also negative values of values are sometimes seen, presumably due to smearing.
Work is continuing on code correctness and execution speed.
Curtis suggested that events with electromagnetic background turned on be generated to measure event size.
David described recent work on generating ROOT trees from HDDM-formaterd data. The idea is to do this automatcially from the information in the template included in every HDDM file, with no hand coding. See his slides for details.
There appears to be a limitation of two levels of nesting in the HDDM structure when conversion to ROOT is attempted. David will study this further.
David discussed recent work implementing a system for getting files (mainly magnetic field maps) that are too large to be handles conveniently by a relational database system (although that is what the CCDB is doing now). Such files are typically called resources. They tend not to change on a run-by-run basis, so caching them on a local disk is an efficient scheme. This is work in progress and we will get a report form David in the future when the package is ready.
Reconstruction sub-group reports
Calorimeters and Tracking
While working on the FDC model in the simulation, Simon noticed some overlaps in the geometry, in particular an overlap between FDC cables and the BCAL and the newly added FDC survey monuments, again with the BCAL. Richard Jones noted some of these in a previous collaboration meeting. David has contacted Zizis and Elton on this and has identified the relevant drawings. A fix is on the way. In addition David noticed that an outer aluminum layer on the BCAL was never included; it will be added as well.
PID: Reorganized Offline HOWTO List
Paul re-organized the wiki page for the software HOWTOs putting them into sensible categories.