<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>After glancing at my notes I found that I missed reporting on a
      couple of items:</p>
    <h3><span class="mw-headline" id="Planning_for_Launches">Planning
        for Launches</span></h3>
    <p>We discussed the idea of doing some semi-formal documentation of
      plans for launches mainly as a communication tool both for the
      collaboration and for computing resource providers. We should at
      least write down things that we know are upcoming acknowledging
      the usual schedule uncertainties. We agreed it was a good idea
      which we should discuss further.
    </p>
    <h3><span class="mw-headline"
        id="Reproducibility_in_Tracking_Reconstruction">Reproducibility
        in Tracking Reconstruction</span></h3>
    <p>Beni thinks he has discovered the problem: somehow duplicate, but
      not identical, FDC pseudo-points are being fed into the tracking.
      Which point gets used can vary from one run of the program to the
      next. The source of the duplication is the next thing to
      investigate. <br>
    </p>
    <p>...and the HDGeant4 meeting will be at 10 am, not at 10 pm as
      originally stated...you can thank Sean for not ruining your Friday
      night.<br>
    </p>
    <br>
    <div class="moz-cite-prefix">On 06/29/2018 03:52 PM, Mark Ito wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:26922efc-c49b-5683-679e-39cbaaee3354@jlab.org">
      <meta http-equiv="content-type" content="text/html; charset=utf-8">
      <p>Folks,</p>
      <p>Please find the minutes below and at</p>
      <p> 
        <a class="moz-txt-link-freetext"
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_June_29,_2018#Minutes"
          moz-do-not-send="true">https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_June_29,_2018#Minutes</a></p>
      <p>  -- Mark</p>
      <p>_______________________</p>
      <p> </p>
      <div id="globalWrapper">
        <div id="column-content">
          <div id="content" class="mw-body" role="main">
            <h2 id="firstHeading" class="firstHeading" lang="en"><span
                dir="auto">GlueX Offline Meeting, June 29, 2018, </span><span
                class="mw-headline" id="Minutes">Minutes</span></h2>
            <div id="bodyContent" class="mw-body-content">
              <div id="mw-content-text" dir="ltr" class="mw-content-ltr"
                lang="en">
                <p>Present: </p>
                <ul>
                  <li> <b> CMU: </b> Naomi Jarvis</li>
                  <li> <b> FIU: </b> Mahmoud Kamel</li>
                  <li> <b> FSU: </b> Sean Dobbs</li>
                  <li> <b> JLab: </b> Alex Austregesilo, Mark Ito
                    (chair), David Lawrence, Simon Taylor, Beni Zihlmann</li>
                  <li> <b> W&M: </b> Justin Stevens</li>
                  <li> <b> Yerevan: </b> Hrach Marukyan</li>
                </ul>
                <p>There is a <a rel="nofollow" class="external text"
                    href="https://bluejeans.com/s/Gl5u7/"
                    moz-do-not-send="true">recording of this meeting</a>
                  on the BlueJeans site. Use your JLab credentials to
                  access it. </p>
                <h3><span class="mw-headline" id="Announcements">Announcements</span></h3>
                <ol>
                  <li> <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2018-June/003228.html"
                      moz-do-not-send="true">Sim-Recon 2.26.0</a>. A
                    periodic release. See the release notes for changes
                    from the last such release.</li>
                  <li> <a rel="nofollow" class="external text"
href="https://github.com/JeffersonLab/gluex_root_analysis/releases/tag/0.3"
                      moz-do-not-send="true">GlueX Root Analysis 0.3</a>.
                    There has not been a new release for about a year.
                    This one was overdue.</li>
                </ol>
                <h3><span class="mw-headline"
                    id="Review_of_minutes_from_the_June_15_meeting">Review
                    of minutes from the June 15 meeting</span></h3>
                <p>We went over <a
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_June_15,_2018#Minutes"
                    title="GlueX Offline Meeting, June 15, 2018"
                    moz-do-not-send="true"> the minutes</a>. </p>
                <ul>
                  <li> The Experimental Readiness Review for GlueX II
                    was held last Monday. David noted that Eugene
                    Chudakov has already circulated the preliminary
                    report of the committee.</li>
                  <li> David mentioned that we may want to revisit our
                    estimate of the amount of Monte Carlo we need. It
                    accounts for nearly half of our current projections.
                    This is in line with what we have always said, but
                    bears re-examination.</li>
                </ul>
                <h3><span class="mw-headline"
                    id="Coming_Computing_Resources">Coming Computing
                    Resources</span></h3>
                <p>Mark flipped through <a rel="nofollow"
                    class="external text"
href="https://www.jlab.org/indico/event/247/session/5/contribution/14/material/slides/0.pdf"
                    moz-do-not-send="true">Chip Watson's talk at the
                    June 28 Round Table</a>. </p>
                <p>Increases to the core count on the farm coming this
                  summer were listed on slide 6: </p>
                <ul>
                  <li> Current system: 3.5k cores (scaled to Broadwell)</li>
                  <li> Major farm upgrade due in July: 88 dual 20 core
                    Skylake compute nodes (farm18), adds 3.5k cores
                    (100% gain)</li>
                  <li> Retiring LQCD cluster to be shared for 6 months,
                    250 dual 8 core (2012 Sandy Bridge) compute nodes,
                    adds 2.4k cores</li>
                  <li> Size for the Fall run: 9.4k cores (up 2.7x)</li>
                  <li> Note: 2.7k cores go end of life mid way through
                    FY2019, and we might </li>
                </ul>
                <p>add only 1.8k new, dropping onsite capacity to 8.5k
                  cores, still up 150%. </p>
                <p>For all of the talks presented at the Round Tables
                  see the <a rel="nofollow" class="external text"
                    href="https://www.jlab.org/indico/event/247/"
                    moz-do-not-send="true">Indico site</a>. </p>
                <h3><span class="mw-headline"
                    id="Report_from_the_June_28_SciComp_Meeting">Report
                    from the June 28 SciComp Meeting</span></h3>
                <p>Mark and David gave the report. </p>
                <ul>
                  <li> SciComp has replaced PBS with <a rel="nofollow"
                      class="external text"
                      href="https://slurm.schedmd.com/"
                      moz-do-not-send="true">Slurm</a> as the underlying
                    batch scheduler for the LQCD cluster and are looking
                    to do the same for the Experimental Nuclear Physics
                    (ENP) cluster. We now have Auger sitting on top of
                    PBS and SWIF sitting on top of Auger. They are
                    looking at eliminating the Auger layer.</li>
                  <li> SciComp is planning on allowing LQCD jobs to run
                    on the ENP cluster at low priority. The reverse will
                    not be allowed. Jobs submitted to ENP will be able
                    to pre-empt LQCD jobs (kill them to make room if no
                    job slots are free).</li>
                </ul>
                <h3><span class="mw-headline"
                    id="Missing_CDC_hits_in_recent_bggen_launch">Missing
                    CDC hits in recent bggen launch</span></h3>
                <p>Alex A. reported on the problem. All CDC hits are
                  apparently missing. Cause of the problem is not known.
                  See his presentation of the evidence starting at 40:40
                  in the <a rel="nofollow" class="external text"
                    href="https://bluejeans.com/s/Gl5u7/"
                    moz-do-not-send="true">recording</a>. </p>
                <p>We will have to re-run. </p>
                <h3><span class="mw-headline"
                    id="Splitting_up_Sim-Recon">Splitting up Sim-Recon</span></h3>
                <p>Mark reported that there was still some remaining
                  work to get to a production system. </p>
                <ol>
                  <li> <b>Environment variable changes</b>. Rather than
                    HALLD_HOME for sim-recon, we will have
                    HALLD_SIM_HOME for the halld_sim repository
                    (originally gluex_sim) and HALLD_RECON_HOME for the
                    halld_recon repository (originally gluex_recon). We
                    decided that the "halld" name reflected the fact
                    that non-GlueX experiments will be using the same
                    software.</li>
                  <li> <b>Preserving branches in halld_recon</b>. There
                    were some details to be worked out to have the
                    branches from sim-recon transmit to halld_recon.</li>
                  <li> <b>Removing sim pieces from gluex_recon</b>.
                    This had not been done. Removal necessitates changes
                    in the build system.</li>
                  <li> <b>Dealing with build_scripts</b>. Build_scripts
                    has not been modified for the new configuration. All
                    work so far has been in getting a working build.</li>
                </ol>
                <h3><span class="mw-headline" id="HDGeant4_Meeting">HDGeant4
                    Meeting</span></h3>
                <p>At the last collaboration meeting Richard suggested
                  having a meeting dedicated to issues related to the
                  Geant4 version of our simulation. The first meeting
                  will be Friday, July 6 at 10 pm. We will likely meet
                  bi-weekly in the slot opposite the Offline Software
                  meetings. </p>
                <h3><span class="mw-headline"
                    id="Review_of_Pull_Requests_and_Software_Help_Topics">Review
                    of Pull Requests and Software Help Topics</span></h3>
                <p>We went over <a rel="nofollow" class="external text"
href="https://github.com/JeffersonLab/sim-recon/pulls?q=is%3Aclosed+is%3Apr"
                    moz-do-not-send="true">recent pull requests</a> and
                  <a rel="nofollow" class="external text"
                    href="https://groups.google.com/forum/#%21forum/gluex-software"
                    moz-do-not-send="true">recent discussion on the
                    GlueX Software Help List</a> without major
                  discussions. </p>
              </div>
            </div>
          </div>
        </div>
      </div>
      <pre class="moz-signature" cols="72">-- 
Mark Ito, <a class="moz-txt-link-abbreviated" href="mailto:marki@jlab.org" moz-do-not-send="true">marki@jlab.org</a>, (757)269-5295
</pre>
      <!--'"--><br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
Halld-offline mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Halld-offline@jlab.org">Halld-offline@jlab.org</a>
<a class="moz-txt-link-freetext" href="https://mailman.jlab.org/mailman/listinfo/halld-offline">https://mailman.jlab.org/mailman/listinfo/halld-offline</a></pre>
    </blockquote>
    <br>
    <pre class="moz-signature" cols="72">-- 
Mark Ito, <a class="moz-txt-link-abbreviated" href="mailto:marki@jlab.org">marki@jlab.org</a>, (757)269-5295
</pre>
  </body>
</html>