GlueX Software Meeting, July 6, 2021
GlueX Software Meeting
Tuesday, July 6, 2021
3:00 pm EDT
BlueJeans: 968 592 007
- Review of Minutes from the Last Software Meeting (all)
- Minutes from the Last HDGeant4 Meeting (all)
- Report from the July 1st SciComp Meeting (Mark)
- Review of recent issues and pull requests:
- Review of recent discussion on the GlueX Software Help List (all)
- Action Item Review (all)
Present: Thomas Britton, Sean Dobbs, Mark Ito (chair), Igal Jaegle, Naomi Jarvis, Simon Taylor, Beni Zihlmann
- Small Cache File Report We went over the web page that Mark sent around last week. Thomas has started moving small files generated by MCwrapper off the cache disk. Sean requested finer-grained information on the location of the problematic files.
- new hdgeant4 ability to generate mu+mu- pairs in GlueX target Richard added this new feature to help with CPP simulations.
- New version of hdgeant4: 2.25.0, new version set: 4.42.0 This new version set contains a new tag of HDGeant4 with this feature.
Review of Minutes from the Last Software Meeting
We went over the minutes from the meeting on June 22nd. There was some significant discussion on action items. Some were dropped. See the list list below for the survivors.
Minutes from the Last HDGeant4 Meeting
We went over the minutes from the meeting on June 29th. On the version upgrade front, Mark has succeeded in building the following combination:
- CentOS 7
- Developer Toolset 8 (gives GCC 8.3.1)
- ROOT 6.24.00
- Geant4 10.06.p01
He is in the process of packaging the build and use instructions.
Report from the July 1st SciComp Meeting
Mark went through the agenda of the meeting. Present at the meeting were Brad Sawatzky, Bryan Hess, Graham Heyes, Markus Diefenthaler, Mark Ito, David Lawrence, Amitoj Singh, Ole Hansen, Nathan Baltzell, and Chris Larrieu. Mark made some remarks about the discussion at the meeting:
- Mark's clean up efforts on /work/halld have stopped with the hope that we have enough margin to make it to release of space on the new work file server. We are at 94% full right now.
- For the last batch of tapes sent to IBM for recovery, uncorrupted files were re-written to tape before the damaged tape was shipped, vastly reducing the number of unavailable files.
- The list of high-priority files missing from the tape library, given to SciComp, is moot at this late stage of recovery. We wait for the normal process to complete.
- There was discussion of Rucio as a possible solution to tracking files generated by data taking and launches. A database of such files could help in deciding if a production effort is done by making it possible inventory the output files produced and determine if any files are missing.
Work Disk Performance
Beni recalled that, some years ago, Alex Austregesilo had trouble merging files using the work disk. The method used at the time generated so much activity that the disk was rendered unusable for all. He wondered if this was a problem that is still with us.
Mark remarked that he is thinking about future guidance on what disk should be used for which purpose. Volatile: staging of large output files; cache: files to be written to tape; work: personal builds and tests. Beni commented that we stopped worrying about work disk space during 2020 because there was plenty of space available.
Action Item Review
- Decide on needed tape volume sets. (Sean, Mark I.)
- Review disk usage when re-repatriating recon launch data. (Alex, Mark)
- Automatically make CCDB entries for selected RCDB items. (Mark I.)
- Modify the documentation on how to copy files from JLab. (Mark I.)
- Generate finer-grained info on small cache files. (Mark I.)