<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>Folks,</p>
    <p>Please find the minutes below and at
<a class="moz-txt-link-freetext" href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_October_4,_2017#Minutes">https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_October_4,_2017#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 Minutes, October 4, 2017</span></h2>
          <h2> </h2>
          <div id="bodyContent" class="mw-body-content">Present:
            <div id="mw-content-text" dir="ltr" class="mw-content-ltr"
              lang="en">
              <ul>
                <li> <b> CMU: </b>: Naomi Jarvis</li>
                <li> <b> FIU </b>: Joerg Reinhold</li>
                <li> <b> FSU </b>: Sean Dobbs</li>
                <li> <b> JLab: </b>: Alex Austregesilo, Amber
                  Boehnlein, Thomas Britton, Eugene Chudakov, Sebastian
                  Cole, Mark Ito (chair), David Lawrence, Sascha Somov,
                  Simon Taylor, Beni Zihlmann</li>
              </ul>
              <p>There is a <a rel="nofollow" class="external text"
                  href="https://bluejeans.com/s/1KDeL/">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/2017-September/002948.html">new
                    tag of mcsmear branch of sim-recon, version 3</a>.
                  Sean released a new tag of the mcsmear development
                  branch of sim-recon. Mark has built it at JLab.</li>
                <li> <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2017-September/002955.html">old,
                    empty volatile directories now being deleted</a>
                  Directories on the volatile disk that are (a) empty
                  and (b) remain so for three months are now being
                  deleted.</li>
                <li> <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2017-September/002954.html">Farm
                    jobs should use SQLite form of the CCDB</a> If
                  hundreds of farm jobs access the MySQL database server
                  for calibration constants, the server gets
                  overwhelmed. Use the disk-resident SQLite version of
                  the database when running on the farm.
                  <ul>
                    <li> Thomas has added an option to MCwrapper for
                      selecting SQLite files for CCDB.</li>
                  </ul>
                </li>
              </ol>
              <h3><span class="mw-headline"
                  id="Review_of_minutes_from_the_last_meeting">Review of
                  minutes from the last meeting</span></h3>
              <p>We went over the <a
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_September_20,_2017#Minutes"
                  title="GlueX Offline Meeting, September 20, 2017">minutes
                  from September 20</a>.
              </p>
              <ul>
                <li> The automatic updates of the <b> Oasis file system
                  </b> are still not functioning fully automatically.
                  Mark has a <a rel="nofollow" class="external text"
                    href="https://ticket.opensciencegrid.org/34965">ticket
                    into the Grid folks</a> asking for guidance on how
                  to proceed.</li>
                <li> The first monthly <a rel="nofollow"
                    class="external text"
href="https://mailman.jlab.org/pipermail/halld-physics/2017-October/001144.html">Analysis
                    Launch is now finished</a>.</li>
              </ul>
              <h3><span class="mw-headline" id="HDvis_update">HDvis
                  update</span></h3>
              <p>Thomas guided us through the latest version of the <a
                  rel="nofollow" class="external text"
                  href="https://halldweb.jlab.org/talks/2017/HDvis2/js/event.html">new
                  event display</a>. See the demo starting at 10:00 in
                the recording.
              </p>
              <ul>
                <li> The orthographic view is now in.</li>
                <li> Thomas has been talking to groups about what they
                  want to see in the display of their detector system.</li>
                <li> He will give a demo at the collaboration meeting.</li>
                <li> David requested that an axis indicator appear.</li>
                <li> Naomi suggested fixed, standard views for
                  re-orienting oneself. Thomas said that that is in the
                  plan.</li>
                <li> Added in press: there was a glitch with mouse-over
                  in the demo. Subsequent to the meeting, Thomas found
                  and fixed it. It appeared only with the Firefox
                  browser.</li>
              </ul>
              <h3><span class="mw-headline"
                  id="Reducing_work_disk_usage">Reducing work disk usage</span></h3>
              <p>Mark went over <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2017-September/002952.html">his
                  recent email</a>. Some points of information:
              </p>
              <ul>
                <li> We are getting a new work disk. It with be ZFS and
                  not Lustre.</li>
                <li> The new disk is significantly smaller than the work
                  space we are using now. We will have to reduce our
                  usage from 77 TB down to 45.</li>
                <li> Some reduction (about 10 TB) has already occurred
                  in response to Mark's email.</li>
                <li> We looked at the <a rel="nofollow" class="external
                    text"
                    href="https://halldweb.jlab.org/disk_management/work_report.html">Work
                    Disk Usage Leader-board</a>.</li>
                <li> Mark reported that it seems that SciComp has been
                  slowing reducing our cache and volatile reservations
                  in response to ever increasing usage on work.
                  <ul>
                    <li> Work is not formally managed. The work
                      allocation is done by UNIX group is really total
                      group Lustre usage minus volatile group usage
                      minus cache group usage. Volatile and cache
                      however are managed by directory location so there
                      are corner cases. As an unmanaged work grows,
                      cache and volatile feel the squeeze.</li>
                  </ul>
                </li>
                <li> Mark will be contacting individuals and consulting
                  on reducing their work disk footprint.</li>
              </ul>
              <p>If you have old, unused files under /work/halld, now
                would be a great time to delete them or archive them to
                tape.
              </p>
              <h3><span class="mw-headline"
                  id="Preliminary_Tape_and_Disk_Usage_Projections">Preliminary
                  Tape and Disk Usage Projections</span></h3>
              <p>Mark went over a <a rel="nofollow" class="external
                  text"
href="https://docs.google.com/spreadsheets/d/1QQQ2R3QrJkJgN37lt9yRlhVylk4KDJQKAn8m6Is8paM/edit?usp=sharing">spreadsheet</a>
                summarizing the files produced by the various launches
                for the Spring 16 and Spring 17 runs. The input was an
                interview with Paul Mattione and Alex on the previous
                day. He outlines two strategies for what we might want
                to keep on disk, high-usage and low-usage. The high
                usage strategy requires 260 TB spinning including both
                Spring 16 and Sprint 17 running. The low-usage only 86
                TB. He noted that our current usage is close to the
                middle of this range.
              </p>
              <p>This is a work in progress. Beni pointed out that Monte
                Carlo data is not counted. Calibrations skims were left
                out also; Mark has not gotten to doing the research on
                those. At this point, these numbers are lower limits.
              </p>
              <h3><span class="mw-headline"
                  id="Review_of_recent_pull_requests">Review of recent
                  pull requests</span></h3>
              <p>Paul checked in a major change to the analysis library.
                He will describe the new stuff at the collaboration
                meeting next week.
              </p>
              <p>A couple of the <a rel="nofollow" class="external
                  text"
href="https://github.com/JeffersonLab/sim-recon/pulls?q=is%3Apr+is%3Aclosed">recent
                  requests</a> (one from David, one from Will McGinley)
                showed large numbers of files differing on the proposed
                branch from the version on the master branch, nearly
                100, when the number of changes being proposed were only
                a few. These were due to Paul's analysis library changes
                getting merged before the requests in question got
                merged. Since there were no conflicts on the two
                branches, these merges can go ahead without causing any
                problem.
              </p>
              <h3><span class="mw-headline"
                  id="Review_of_recent_discussion_on_the_GlueX_Software_Help_List">Review
                  of recent discussion on the GlueX Software Help List</span></h3>
              <p>We went over the <a rel="nofollow" class="external
                  text"
                  href="https://groups.google.com/forum/#%21forum/gluex-software">recent
                  threads</a>.
              </p>
              <p>As the result of <a rel="nofollow" class="external
                  text"
href="https://groups.google.com/forum/#%21topic/gluex-software/2FqnXJepUY8">one
                  discussion</a> Mark built a version of the code from
                the last reconstruction launch on Spring 16 data. In the
                future, we may want to do builds of certain launches on
                the group disk as a matter of course.
              </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">marki@jlab.org</a>, (757)269-5295
</pre>
  </body>
</html>