[Halld-offline] Software Meeting Minutes, January 7, 2021

[Mark Ito]

Folks,

Please find the minutes here 
<https://halldweb.jlab.org/wiki/index.php/GlueX_Software_Meeting,_January_5,_2021#Minutes> 
and below.

   -- Mark

_______________________________


    GlueX Software Meeting, January 5, 2021, Minutes

Present: Alex Austregesilo, Thomas Britton, Sean Dobbs, Sergey Furletov, 
Mark Ito (chair), Naomi Jarvis, David Lawrence, Simon Taylor, Nilanga 
Wickramaarachchi

There is a recording of this meeting 
<https://bluejeans.com/s/6v1iVGDMzH9/>. Log into the BlueJeans site 
<https://jlab.bluejeans.com> first to gain access (use your JLab 
credentials).


      Announcements

 1. New version set: version_4.33.0.xml
    <https://mailman.jlab.org/pipermail/halld-offline/2020-December/008404.html>.
    An version set with update releases of halld_recon, halld_sim, and
    hdgeant4.
 2. New version of halld_sim, 4.26.0, and corresponding recon version
    sets
    <https://mailman.jlab.org/pipermail/halld-offline/2020-December/008416.html>.
    An update release of halld_sim with changes to the MC_GEN event
    generator.
 3. Tape Library Problem Status
    <https://halldweb.jlab.org/talks/2021/tape_situation.pdf> Mark went
    over information he received from Bryan Hess before the break.
    Please see the link for a description of the status at that time.


      Review of Minutes from the Last Software Meeting

We went over the minutes from December 8 
<https://halldweb.jlab.org/wiki/index.php/GlueX_Software_Meeting,_December_8,_2020#Minutes> 
without significant comment.


      Minutes from the Last HDGeant4 Meeting

We went over the minutes from December 15 
<https://halldweb.jlab.org/wiki/index.php/HDGeant4_Meeting,_December_15,_2020#Minutes> 
without significant comment.


      Report from Meeting on Reconstruction on the Grid

Mark gave a summary of the meeting held on December 21 to discuss doing 
reconstruction on the OSG. In attendance at that meeting were Nathan 
Baltzell, Thomas Britton, Edgar Farjado, Bryan Hess, Igal Jaegle, 
Richard Jones, Zisis Papandreou, Maurizio Ungaro, and MMI. Most of the 
meeting was taken up with Richard explaining the scheme he is using for 
a pilot system he is running out of UConn. [Added in press: Richard gave 
a talk at the January 6 Production and Analysis meeting that described 
the system. Please see this talk 
<https://drive.google.com/file/d/126H_isWqHrzZzErORmpmCqtj4i1HFd3H/view?usp=sharing> 
for details.

Richard will be giving a talk tomorrow at a meeting that Edgar attends 
with OSG and the XROOTD team. He will explain his system and solicit 
feedback from those groups.


      Issue with reconstruction test

Alex introduced a problem he has seen where the reconstructions tests 
are running very slowly, as if only one working thread is running when 
many are requested. He showed a plot of the historical processing rate 
<https://halldweb.jlab.org/wiki/images/d/dd/Processing_speed_2021.PNG> 
of the jobs going back about four years. He is referring to the fall-off 
seen in the last few iterations.

[Added in press: this effect may be due to a change in the nightly 
builds that Mark put in. He had started building with profiling 
information included (the -p switch). That has now been removed. We will 
see if the reversion to the old configuration restores the processing 
rate.]


      Review of recent issues and pull requests

Richard has been working on reproducibility issues when running the 
standard reconstruction job. See Issue #492: Repeated runs not giving 
the same results <https://github.com/JeffersonLab/halld_recon/issues/492>.


      Action Items

 1. Ask the Computer Center about staging the start of jobs in time. (Mark)
 2. Make sure that the automatic tests of HDGeant4 pull requests have
    been fully implemented. (Mark, Sean)
 3. Finish conversion of halld_recon to use JANA2. (Nathan)
 4. Ping Dmitry on CCDB Issue #68. (Mark)
 5. Remind Ashley about her problematic cache files. (Mark)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20210107/557d8e3b/attachment.html>