Difference between revisions of "GlueX Software Meeting, December 10, 2019"

From GlueXWiki
Jump to: navigation, search
(Agenda)
(Added the minutes.)
Line 33: Line 33:
  
 
Talks can be deposited in the directory <code>/group/halld/www/halldweb/html/talks/2019</code> on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2019/ .
 
Talks can be deposited in the directory <code>/group/halld/www/halldweb/html/talks/2019</code> on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2019/ .
 +
 +
== Minutes ==
 +
 +
Present:
 +
 +
* ''' JLab: ''' Alex Austregesilo, Thomas Britton, Mark Dalton, Mark Ito (chair), Igal Jaegle, Naomi Jarvis, David Lawrence, Keigo Mizutani, Justin Stevens, Simon Taylor, Beni Zihlmann
 +
* ''' FSU: ''' Sean Dobbs
 +
 +
There is a [https://bluejeans.com/s/60@6B/ recording of his meeting] on the BlueJeans site. Use your JLab credentials to access it.
 +
 +
=== Announcements ===
 +
 +
# From the Computer Center: [https://mailman.jlab.org/pipermail/halld-offline/2019-November/003825.html CentOS 7.7 update and post-Thanksgiving plans]: New AMD farm nodes and more Lustre space.
 +
# [https://mailman.jlab.org/pipermail/halld-offline/2019-December/003832.html GlueX Software build for JLab CentOS 7.7]: CentOS7.7 treated as a separate platform, complete build on it is available.
 +
# [https://github.com/JeffersonLab/gluex_MCwrapper/releases/tag/v2.3.0 MCwrapper 2.3.0 (Frost Release)] Thomas released this today. It contains many, accumulated improvement, most of them under the hood, include multi-process monitoring of jobs. See the release notes for details.
 +
 +
=== Conversion from Python 2 to Python 3 ===
 +
 +
Beni raised the issue. He is concerned that future distributions of Linux will not support Python 2. We may have to react to this as early as January 2020, when support of Python 2 will be withdrawn. Mark thought that the problem may not be all that pressing. He advocates conversion as we identify problems maintaining critical portions of our Python 2 code base.
 +
 +
=== Review of Minutes from the November 26 Meeting ===
 +
 +
We went over [[GlueX Software Meeting, November 26, 2019#Minutes|the minutes]] without significant comment.
 +
 +
=== Review of Minutes from the December 3 Meeting ===
 +
 +
We went over [[HDGeant4 Meeting, December 3, 2019#Minutes|the minutes]] without significant comment.
 +
 +
=== New Version Set and Build ===
 +
 +
Beni requested a new version set. The current default will not support analysis of the data currently being obtained. Mark I. put it on his list.
 +
 +
=== CCDB 2.0 On the Way ===
 +
 +
Mark reported that Dmitry is nearly ready to release CCDB 2.0. It will have improvements including the ability to freeze variation ancestors with letting the variation of interest move forward in time. This can be done without in a way that is fully backward compatible: old code and new code can use the same updated schema. The greatly simplifies implementation in comparison with what one might have expected for a schema change.
 +
 +
=== RCDB Errors in the Counting Room ===
 +
 +
Naomi and Justin reported [https://logbooks.jlab.org/entry/3748503 problems with RCDB information] being recorded as we take data. Naomi will contact Sergey Furletov or Dmitry Romanov or both in an as-yet-to-be-specified order about the problem.
 +
 +
=== MySQL Server Problems: hallddb-farm.jlab.org ===
 +
 +
hallddb-farm is the recommended server to use on the JLab farm for accessing CCDB data. [https://groups.google.com/forum/#!topic/gluex-software/eG6cmxZxyTM Kevin Luckas and Justin have reported jobs hanging forever] when using this server, both from within batch jobs and interactively. Mark I. started looking into it but ran into other, unrelated problems. This needs to be addressed soon!
 +
 +
=== genBH and Non-Standard HDDM Output Data ===
 +
 +
Mark D. reported problems with genBH, an optionally built event generator from the HDGeant4 repository. It produces output in the hddm_mc_s schema of HDDM and not in the usual hddm_s schema. Programs downstream of event generation cannot grok these events. He has modified a private version to use hddm_s. The package requires Richard's Dirac++ package, which we do not build as part of the standard software stack. There was discussion, not conclusive, about whether the modified version should be put into some other repository, presumably halld_sim.
 +
 +
=== Action Item Review ===
 +
 +
# Create and build a new version set. (Mark)
 +
# Test and debug hallddb-farm.jlab.org.
 +
# Figure out what to do with genBH.

Revision as of 21:51, 10 December 2019

GlueX Software Meeting
Tuesday, December 10, 2019
2:00 pm EST
JLab: CEBAF Center A110
BlueJeans: 968 592 007

Agenda

  1. Announcements
    1. Today meeting room booked only until 3:00 PM EST.
    2. From the Computer Center: CentOS 7.7 update and post-Thanksgiving plans
    3. GlueX Software build for JLab CentOS 7.7 (Mark)
    4. MCwrapper 2.3.0 (Frost Release) (Thomas)
  2. Review of minutes from the last Software Meeting (all)
  3. Review of minutes from the last HDGeant4 Meeting (all)
  4. Review of recent issues and pull requests:
    1. halld_recon
    2. halld_sim
    3. CCDB
    4. RCDB
  5. Review of recent discussion on the GlueX Software Help List (all)
  6. Action Item Review (all)

Slides

Talks can be deposited in the directory /group/halld/www/halldweb/html/talks/2019 on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2019/ .

Minutes

Present:

  • JLab: Alex Austregesilo, Thomas Britton, Mark Dalton, Mark Ito (chair), Igal Jaegle, Naomi Jarvis, David Lawrence, Keigo Mizutani, Justin Stevens, Simon Taylor, Beni Zihlmann
  • FSU: Sean Dobbs

There is a recording of his meeting on the BlueJeans site. Use your JLab credentials to access it.

Announcements

  1. From the Computer Center: CentOS 7.7 update and post-Thanksgiving plans: New AMD farm nodes and more Lustre space.
  2. GlueX Software build for JLab CentOS 7.7: CentOS7.7 treated as a separate platform, complete build on it is available.
  3. MCwrapper 2.3.0 (Frost Release) Thomas released this today. It contains many, accumulated improvement, most of them under the hood, include multi-process monitoring of jobs. See the release notes for details.

Conversion from Python 2 to Python 3

Beni raised the issue. He is concerned that future distributions of Linux will not support Python 2. We may have to react to this as early as January 2020, when support of Python 2 will be withdrawn. Mark thought that the problem may not be all that pressing. He advocates conversion as we identify problems maintaining critical portions of our Python 2 code base.

Review of Minutes from the November 26 Meeting

We went over the minutes without significant comment.

Review of Minutes from the December 3 Meeting

We went over the minutes without significant comment.

New Version Set and Build

Beni requested a new version set. The current default will not support analysis of the data currently being obtained. Mark I. put it on his list.

CCDB 2.0 On the Way

Mark reported that Dmitry is nearly ready to release CCDB 2.0. It will have improvements including the ability to freeze variation ancestors with letting the variation of interest move forward in time. This can be done without in a way that is fully backward compatible: old code and new code can use the same updated schema. The greatly simplifies implementation in comparison with what one might have expected for a schema change.

RCDB Errors in the Counting Room

Naomi and Justin reported problems with RCDB information being recorded as we take data. Naomi will contact Sergey Furletov or Dmitry Romanov or both in an as-yet-to-be-specified order about the problem.

MySQL Server Problems: hallddb-farm.jlab.org

hallddb-farm is the recommended server to use on the JLab farm for accessing CCDB data. Kevin Luckas and Justin have reported jobs hanging forever when using this server, both from within batch jobs and interactively. Mark I. started looking into it but ran into other, unrelated problems. This needs to be addressed soon!

genBH and Non-Standard HDDM Output Data

Mark D. reported problems with genBH, an optionally built event generator from the HDGeant4 repository. It produces output in the hddm_mc_s schema of HDDM and not in the usual hddm_s schema. Programs downstream of event generation cannot grok these events. He has modified a private version to use hddm_s. The package requires Richard's Dirac++ package, which we do not build as part of the standard software stack. There was discussion, not conclusive, about whether the modified version should be put into some other repository, presumably halld_sim.

Action Item Review

  1. Create and build a new version set. (Mark)
  2. Test and debug hallddb-farm.jlab.org.
  3. Figure out what to do with genBH.