<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    Folks, find the minutes below and at
    <a class="moz-txt-link-freetext" href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_June_8,_2016#Minutes">https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_June_8,_2016#Minutes</a>
    .<br>
    <br>
      -- Mark<br>
    ___________________________<br>
    <br>
    <div id="globalWrapper">
      <div id="column-content">
        <div id="content" class="mw-body" role="main">
          <h1 id="firstHeading" class="firstHeading" lang="en"><span
              dir="auto">GlueX Offline Meeting, June 8, 2016, </span><span
              class="mw-headline" id="Minutes">Minutes</span></h1>
          <div id="bodyContent" class="mw-body-content">
            <div id="mw-content-text" dir="ltr" class="mw-content-ltr"
              lang="en">
              <p>You can <a rel="nofollow" class="external text"
                  href="https://bluejeans.com/s/9NeN/">view a recording
                  of this meeting</a> on the BlueJeans site.
              </p>
              <p>Present:
              </p>
              <ul>
                <li> <b>FSU</b>: Alyssa Henderson</li>
                <li> <b>IU</b>: Matt Shepherd</li>
                <li> <b>JLab</b>: Alexander Austregesilo, Mark Ito
                  (chair), David Lawrence, Paul Mattione, Nathan Sparks,
                  Justin Stevens</li>
                <li> <b>NU</b>: Sean Dobbs</li>
              </ul>
              <h3><span class="mw-headline" id="Announcements">Announcements</span></h3>
              <ul>
                <li> <b>Nightly Build Re-Config</b>. Mark sent out <a
                    rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-June/002382.html">an
                    email describing the changes due to the change to
                    GCC 4.8 or better</a>. The <a
href="https://halldweb.jlab.org/wiki/index.php/Nightly_Builds_of_GlueX_Software"
                    title="Nightly Builds of GlueX Software">line-up of
                    nodes where the build is done</a> has been updated
                  on the wiki.</li>
                <li> <b>Small Files on Cache Disk</b>. Mark went over <a
                    rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-June/002383.html">his
                    recent email announcing a program of deleting small
                    files from the cache disk</a>.
                  <ul>
                    <li> Sean has already started clean-up of files
                      produced by production runs so far. Accounting
                      files have been removed; ROOT files have been
                      combined resulting in one file per run.</li>
                    <li> The reconstruction launch is producing skims of
                      synch events that are very small and are being
                      written to cache. For now they will not be written
                      to tape; we have some time to decide whether to
                      move them or let them be written to tape from
                      their current location.</li>
                  </ul>
                </li>
                <li> <b>Dropping selected tags from EVIO data</b>. Sean
                  has checked in a change that allows this. He estimates
                  that the reduction of the size of the π<sup>0</sup>
                  skims is an order of magnitude. This method has not
                  been deployed yet. He is waiting for feedback from the
                  calorimeter calibrators.</li>
              </ul>
              <h3><span class="mw-headline"
                  id="Review_of_minutes_from_May_25">Review of minutes
                  from May 25</span></h3>
              <p>We went over <a
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_May_25,_2016#Minutes"
                  title="GlueX Offline Meeting, May 25, 2016">the
                  minutes</a>. Mark mentioned that he sent in <a
                  rel="nofollow" class="external text"
href="https://groups.google.com/forum/#%21topic/gluex-software/c-N3pHwPGXk">the
                  request to Sandy Philpott for the 200 TB of Lustre
                  space</a> we think we will need.
              </p>
              <h3><span class="mw-headline"
                  id="Spring_2016_Run_Processing_Status">Spring 2016 Run
                  Processing Status</span></h3>
              <p>Paul submitted the jobs for the reconstruction launch
                this morning.
              </p>
              <ul>
                <li> The first 50 jobs failed due to SWIF errors.
                  Subsequent jobs started successfully. There are 200
                  jobs running now.</li>
                <li> There are 6500 jobs altogether. Each job requests
                  14 cores.</li>
                <li> There is no longer a mode where we can request an
                  entire node in a way that is independent of the number
                  of cores on a node. We should look at whether
                  restoring that capability should be restored.</li>
                <li> Paul has tagged releases of sim-recon and hdds used
                  for this launch: recon-2016_02-ver01
                  <ul>
                    <li> Mark mentioned he has tagged sim-recon and hdds
                      as well using the numeric scheme. Justin suggested
                      those tags be brought into alignment with the
                      launch tags.</li>
                  </ul>
                </li>
              </ul>
              <h3><span class="mw-headline" id="simX.X">simX.X</span></h3>
              <h4><span class="mw-headline" id="sim1.1_Conditions">sim1.1
                  Conditions</span></h4>
              <p>Sean has put together <a rel="nofollow"
                  class="external text"
href="https://github.com/JeffersonLab/gluex_simulations/tree/master/sim1.1">sim1.1
                  a set of conditions</a> for sim1.1. The plan is to run
                similar statistics to sim1. This data set will be used
                as input for Paul's efficiency and resolution plug-ins
                and the results will be compared with real data to
                obtain corrections for mcsmear. He has identified the
                collimator size (3.4 mm or 5 mm) as a condition that
                should be simulated separately.
              </p>
              <p>He proposes using real run numbers and distributing the
                simulated data in proportion to the distribution in real
                data.
              </p>
              <p>Matt also suggested we think about simulating signals
                with appropriate run dependent conditions.
              </p>
              <h4><span class="mw-headline"
                  id="sim2.0--Efficiency_Matching">sim2.0--Efficiency
                  Matching</span></h4>
              <p>Sean proposes another set of simulation, with higher
                statistics, incorporating what we learn from sim1.1. As
                part of this[?] he would like to see us institute some
                form of channel-by-channel, hit level efficiencies. He
                has already started <a
href="https://halldweb.jlab.org/wiki/index.php/Mcsmear_updates#Per-Channel_Efficiency"
                  title="Mcsmear updates">designing CCDB tables to
                  institute such a scheme</a>.
              </p>
              <p>Paul cautioned us that faithful simulation of detector
                efficiencies might be more complex than independent
                channel-by-channel random number generation. We agreed
                that that was generally the case, but this scheme
                provides a good starting point.
              </p>
              <h3><span class="mw-headline"
                  id="Access_to_the_RCDB_during_reconstruction_and_analysis">Access
                  to the RCDB during reconstruction and analysis</span></h3>
              <p>Mark is working on a scheme to add RCDB to the standard
                software suite so that we can take advantage of Dmitry
                Romanov's C++ API to the RCDB.
              </p>
              <p>Mark reminded us that the RCDB does not have a history
                mechanism. It is not possible to set a time-stamp to
                recover a previous incarnation of what the RCDB thought
                the running conditions were. Previous versions are only
                available via backups. We might want to enhance our
                backup system to provide for easier user access to past
                versions of the database.
              </p>
              <h3><span class="mw-headline" id="DTreeInterface">DTreeInterface</span></h3>
              <p>Paul took us through <a rel="nofollow" class="external
                  text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-June/002385.html">his
                  email announcing a new interface</a> for filling ROOT
                trees where best-practices locking is done transparently
                for the user. Users do not deal with ROOT objects
                directly; all access is done through the interface. One
                caution is that the user has to make sure to destroy the
                interface object at the end of the job so the interface
                knows when to close the ROOT file.
              </p>
              <p>There is an incompatibility with the Apple-provided
                version of the Clang compiler. Nathan has finessed this
                in the HDPM system.
              </p>
            </div>
            <div class="printfooter">
              Retrieved from "<a dir="ltr"
href="https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_June_8,_2016&oldid=75548">https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_June_8,_2016&oldid=75548</a>"</div>
          </div>
        </div>
      </div>
      <div id="footer" role="contentinfo">
        <ul id="f-list">
          <li id="lastmod"> This page was last modified on 9 June 2016,
            at 10:12.</li>
        </ul>
      </div>
    </div>
    <br>
    <pre class="moz-signature" cols="72">-- 
<a class="moz-txt-link-abbreviated" href="mailto:marki@jlab.org">marki@jlab.org</a>, (757)269-5295
</pre>
  </body>
</html>