[Halld-offline] Offline Software Meeting Minutes, August 9, 2017

[Mark Ito]

Please find the minutes below and at 
https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_August_9,_2017#Minutes 
.

____________


    Minutes

Present:

  * *FSU *: Sean Dobbs
  * *Glasgow *: Peter Pauli
  * *JLab *: Alexander Austregesilo, Amber Boehnlein, Thomas Britton,
    Eugene Chudakov, Hovanes Egiyan, Mark Ito (chair), Curtis Meyer,
    Sandy Philpott, Joerg Reinhold, Dmitry Romanov, Simon Taylor, Beni
    Zihlmann
  * *MIT *: Cris Fanelli

There is a recording of this meeting <https://bluejeans.com/s/w6yEV/> on 
the BlueJeans site. Use your JLab credentials to access it.


      Announcements


        Analysis Launch

Alex announced that the analysis launch is done 
<https://mailman.jlab.org/pipermail/halld-physics/2017-August/001072.html>. 
It took about three days. Sixty channels were analyzed, creating 2.5 
million ROOT trees. In the next attempt these trees will have to be 
consolidated into fewer, bigger files so that they can be written to 
tape more easily. The current set of files caused a backup of the tape 
library. Trees are produced one per file. The plan is to merge all trees 
from a given channel in a given run.


        Mini-Launches of Simulation on the OSG

Sean has been using the infrastructure provided by Richard Jones to run 
simulation jobs on the OSG. Jobs were submitted from the new JLab Submit 
Host. Up to 2000 jobs were running at a time.

Sean wants to work on the framework so that jobs can be sent either to 
the OSG or the JLab farm.


        Collaboration Date Moved

Curtis announced the that next collaboration meeting will be held 
October 12,13, and 14.


      Review of Minutes from the Last Meeting

We reviewed the minutes from the meeting on July 26 
<https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_July_26,_2017#Minutes>. 


  * Sandy reported that the move of central computing nodes from
    upstairs at CEBAF Center to the Data Center on July 29 went well.
  * Thomas reported that he is currently working on aligning FDC hits in
    the event display, working closely with Simon. The BCAL hits were
    added last week. Dmitry has been working on the framework,
    organizing the existing code. To begin with, they are recommending a
    mode where users save events of interest during reconstruction for
    viewing at a later stage rather than navigating through events as
    reconstruction is performed.


      Coverity

The folks at Coverity 
<https://www.synopsys.com/software-integrity/security-testing/static-analysis-sast.html> 
have analyzed code from Hall B and Hall D. They gave a presentation last 
week showing results. The program does a static analysis of code, 
flagging code patterns that may indicate errors, reporting them to the 
user with a nice graphical interface.

There seems to be general agreement that the cost of a license could pay 
big dividends by avoiding future debugging effort by code developers. 
Amber mentioned that the Lab would likely purchase a license for a small 
number of seats for an initial trial period.


      Running ROOT in Debug Mode

Beni described a method for getting source code information from the 
stack trace generated when errors are encountered running compiled ROOT 
macros. When compiling macros, append "++g", for example:

  .L MyMacro.C++g

See Beni for all of the details.

-- 
Mark Ito, marki at jlab.org, (757)269-5295

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20170810/cbb27eac/attachment.html>