<html>
  <head>
    <meta content="text/html; charset=windows-1252"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <p>I think there is an api in the multi-block writer to disable
      writing metadata. Utkarsh knows, he's the one that figured it out.
      not sure how it applies to your use case. you may not even be
      using multi-block format...<br>
    </p>
    <br>
    <div class="moz-cite-prefix">On 05/23/2016 06:57 AM, Gallagher,
      Timothy P wrote:<br>
    </div>
    <blockquote
cite="mid:BN3PR07MB2627F93DC800BA34D03EF05F9D4E0@BN3PR07MB2627.namprd07.prod.outlook.com"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=windows-1252">
      <style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;} --></style>
      <div id="divtagdefaultwrapper"
style="font-size:12pt;color:#000000;background-color:#FFFFFF;font-family:Calibri,Arial,Helvetica,sans-serif;">
        <p>Interesting -- I'd be willing to try it out if you have more
          details (commit hashtag, ticket, etc.), even if I have to
          figure out how to backport it from a newer version.
          <br>
        </p>
        <p><br>
        </p>
        <p>Thanks,</p>
        <p><br>
        </p>
        <p>Tim<br>
        </p>
        <br>
        <br>
        <div style="color: rgb(49, 55, 57);">
          <hr tabindex="-1" style="display:inline-block; width:98%">
          <div id="divRplyFwdMsg" dir="ltr"><font style="font-size:11pt"
              color="#000000" face="Calibri, sans-serif"><b>From:</b>
              Burlen Loring <a class="moz-txt-link-rfc2396E" href="mailto:burlen.loring@gmail.com"><burlen.loring@gmail.com></a><br>
              <b>Sent:</b> Sunday, May 22, 2016 10:30 PM<br>
              <b>To:</b> Gallagher, Timothy P; Andy Bauer<br>
              <b>Cc:</b> <a class="moz-txt-link-abbreviated" href="mailto:paraview@paraview.org">paraview@paraview.org</a><br>
              <b>Subject:</b> Re: [Paraview] Memory leak with Catalyst</font>
            <div> </div>
          </div>
          <div>btw, we just had similar problem that was entirely
            explained by vtk writer attempting to gather a bunch of
            arrays to the root node. Utkarsh found a way to disable that
            behavior. I wonder if you are hitting the same.
            <br>
            <br>
            <div class="moz-cite-prefix">On 5/20/2016 4:36 PM,
              Gallagher, Timothy P wrote:<br>
            </div>
            <blockquote type="cite">
              <div id="divtagdefaultwrapper" style="font-size:12pt;
                color:#000000; background-color:#FFFFFF;
                font-family:Calibri,Arial,Helvetica,sans-serif">
                <p>Well... not going so well. <br>
                </p>
                <p><br>
                </p>
                <p>If I run a small simulation with massif and the
                  Release build of paraview, I don't see any growing
                  memory in time. I also get my VTK files and images.
                  <br>
                </p>
                <p><br>
                </p>
                <p>If I run the same simulation linked against a Debug
                  build with VTK_DEBUG_LEAKS on, there is a segfault
                  inside one of the libvtkRendering libraries (it says
                  libvtkRenderingOp but the rest of the name is cut
                  off). If I take the image rendering out of my pipeline
                  (comment out the RegisterView call) and leave the VTK
                  file writing, it runs fine and doesn't report any
                  leaks. I tried to get the segfault to drop a core
                  file, but it won't do it. </p>
                <p><br>
                </p>
                <p>I'm thinking that there is an issue in the rendering
                  part causing the leak. I have no idea what it could be
                  though -- this is the 4.4.0 version, is anybody aware
                  of any bug fixes between 4.4.0 and current versions
                  that might be relevant?
                  <br>
                </p>
                <p><br>
                </p>
                <p>Thanks for the help,</p>
                <p><br>
                </p>
                <p>Tim<br>
                </p>
                <p><br>
                </p>
                <p><br>
                </p>
                <br>
                <br>
                <div style="color:rgb(0,0,0)">
                  <hr tabindex="-1" style="display:inline-block;
                    width:98%">
                  <div id="divRplyFwdMsg" dir="ltr"><font
                      style="font-size:11pt" color="#000000"
                      face="Calibri, sans-serif"><b>From:</b> ParaView
                      <a moz-do-not-send="true"
                        class="moz-txt-link-rfc2396E"
                        href="mailto:paraview-bounces@paraview.org"><paraview-bounces@paraview.org></a>
                      on behalf of Gallagher, Timothy P
                      <a moz-do-not-send="true"
                        class="moz-txt-link-rfc2396E"
                        href="mailto:tim.gallagher@gatech.edu"><tim.gallagher@gatech.edu></a><br>
                      <b>Sent:</b> Friday, May 20, 2016 4:17 PM<br>
                      <b>To:</b> Burlen Loring; Andy Bauer<br>
                      <b>Cc:</b> <a moz-do-not-send="true"
                        class="moz-txt-link-abbreviated"
                        href="mailto:paraview@paraview.org">
                        paraview@paraview.org</a><br>
                      <b>Subject:</b> Re: [Paraview] Memory leak with
                      Catalyst</font>
                    <div> </div>
                  </div>
                  <div>
                    <div id="divtagdefaultwrapper"
                      style="font-size:12pt; color:#000000;
                      background-color:#FFFFFF;
                      font-family:Calibri,Arial,Helvetica,sans-serif">
                      <p>Thanks for the advice Burlen -- I haven't used
                        Massif before, so this is a good opportunity to
                        give it a try.
                        <br>
                      </p>
                      <p><br>
                      </p>
                      <p>The initialize/finalize each time approach
                        doesn't work either. I am calling the Finalize()
                        routine on the vtkCPProcessor class but then
                        when it tries to initialize again, it throws:</p>
                      <p><br>
                      </p>
                      <div>Warning: In
                        /data4/ParaView/VTK/Parallel/Core/vtkSocketController.cxx,
                        line 50<br>
                        vtkSocketController (0xa4342f0): Already
                        initialized.<br>
                      </div>
                      <p><br>
                      </p>
                      <p>and then crashes. So, hopefully I can sort out
                        the memory leak rather than trying to debug why
                        a hacked fix doesn't work!</p>
                      <p><br>
                      </p>
                      <p>Tim<br>
                      </p>
                      <br>
                      <br>
                      <div style="color:rgb(49,55,57)">
                        <hr tabindex="-1" style="display:inline-block;
                          width:98%">
                        <div id="divRplyFwdMsg" dir="ltr"><font
                            style="font-size:11pt" color="#000000"
                            face="Calibri, sans-serif"><b>From:</b>
                            Burlen Loring
                            <a moz-do-not-send="true"
                              class="moz-txt-link-rfc2396E"
                              href="mailto:burlen.loring@gmail.com"><burlen.loring@gmail.com></a><br>
                            <b>Sent:</b> Friday, May 20, 2016 3:46 PM<br>
                            <b>To:</b> Gallagher, Timothy P; Andy Bauer<br>
                            <b>Cc:</b> <a moz-do-not-send="true"
                              class="moz-txt-link-abbreviated"
                              href="mailto:paraview@paraview.org">
                              paraview@paraview.org</a><br>
                            <b>Subject:</b> Re: [Paraview] Memory leak
                            with Catalyst</font>
                          <div> </div>
                        </div>
                        <div>
                          <p>VTK_DEBUG_LEAKS, although will catch some
                            serious errors, will not catch all kinds of
                            leaks. For example you can have leaks where
                            data erroneously accumulates with each time
                            step, but is deleted at program termination.
                            VTK_DEBUG_LEAKS will not catch that kind of
                            error. It's probably better to use massif to
                            profile your code on a small one node run.
                            There's an excellent tool called massif
                            visualizer to aid in exploring the data
                            generated.</p>
                          <p><br>
                          </p>
                          <p>$0.02<br>
                          </p>
                          <br>
                          <div class="moz-cite-prefix">On 05/20/2016
                            11:56 AM, Gallagher, Timothy P wrote:<br>
                          </div>
                          <blockquote type="cite">
                            <div id="divtagdefaultwrapper"
                              style="font-size:12pt; color:#000000;
                              background-color:#FFFFFF;
                              font-family:Calibri,Arial,Helvetica,sans-serif">
                              <p>Hi Andy,</p>
                              <p><br>
                              </p>
                              <p>Thanks for the tips. I will get working
                                on the VTK_DEBUG_LEAKS now and see what
                                it turns up.
                                <br>
                              </p>
                              <p><br>
                              </p>
                              <p>The initialize/finalize every time is
                                definitely a hack and not for long-term
                                use. But, sponsors want a report on
                                Monday and we need to be able to
                                visualize things for that -- the
                                simulation is too big to write data
                                files and post-process later. So I
                                modified the code to do that for now
                                until I can find a proper fix. <br>
                              </p>
                              <p><br>
                              </p>
                              <p>I'll let you know if I get stuck with
                                the log file. <br>
                              </p>
                              <p><br>
                              </p>
                              <p>Thanks again,</p>
                              <p><br>
                              </p>
                              <p>Tim<br>
                              </p>
                              <br>
                              <br>
                              <div style="color:rgb(49,55,57)">
                                <hr tabindex="-1"
                                  style="display:inline-block;
                                  width:98%">
                                <div id="divRplyFwdMsg" dir="ltr"><font
                                    style="font-size:11pt"
                                    color="#000000" face="Calibri,
                                    sans-serif"><b>From:</b> Andy Bauer
                                    <a moz-do-not-send="true"
                                      class="moz-txt-link-rfc2396E"
                                      href="mailto:andy.bauer@kitware.com"><andy.bauer@kitware.com></a><br>
                                    <b>Sent:</b> Friday, May 20, 2016
                                    2:39 PM<br>
                                    <b>To:</b> Gallagher, Timothy P<br>
                                    <b>Cc:</b> <a
                                      moz-do-not-send="true"
                                      class="moz-txt-link-abbreviated"
                                      href="mailto:paraview@paraview.org">
                                      <a class="moz-txt-link-abbreviated" href="mailto:paraview@paraview.org">paraview@paraview.org</a></a><br>
                                    <b>Subject:</b> Re: [Paraview]
                                    Memory leak with Catalyst</font>
                                  <div> </div>
                                </div>
                                <div>
                                  <div dir="ltr">
                                    <div>
                                      <div>
                                        <div>
                                          <div>Hi Tim,<br>
                                            <br>
                                          </div>
                                          If you build Catalyst with
                                          VTK_DEBUG_LEAKS enabled it is
                                          pretty good at finding VTK
                                          objects that aren't deleted
                                          properly. You should be able
                                          to run this with a small
                                          amount of calls to Catalyst as
                                          well. If you try this and want
                                          help understanding the output
                                          (if an object like a
                                          vtkUnstructuredGrid is leaked
                                          it will often give a whole
                                          bunch of false leaks that the
                                          unstructured grid is just
                                          holding the references to),
                                          just share your output file
                                          and I can take a look at it to
                                          try and narrow down the
                                          culprit. That may be slightly
                                          easier to use than Valgrind.<br>
                                          <br>
                                        </div>
                                        <div>Beyond this, you could
                                          maybe try the same run but
                                          without doing any Catalyst
                                          work to see what happens then.
                                          That may be a lot of compute
                                          cycles but I'm not sure how
                                          else to try and bisect where
                                          the memory leak is occurring.<br>
                                        </div>
                                        <div><br>
                                        </div>
                                        Initializing and finalizing
                                        Catalyst every time you want
                                        output would probably work but
                                        I'd think there may be
                                        significant overhead doing it
                                        like this. Plus, it's not really
                                        solving the problem -- just
                                        avoiding it.<br>
                                        <br>
                                      </div>
                                      Best,<br>
                                    </div>
                                    Andy<br>
                                    <div>
                                      <div>
                                        <div><br>
                                        </div>
                                      </div>
                                    </div>
                                  </div>
                                  <div class="gmail_extra"><br>
                                    <div class="gmail_quote">On Fri, May
                                      20, 2016 at 12:57 PM, Gallagher,
                                      Timothy P <span dir="ltr">
                                        <<a moz-do-not-send="true"
                                          href="mailto:tim.gallagher@gatech.edu"
                                          target="_blank">tim.gallagher@gatech.edu</a>></span>
                                      wrote:<br>
                                      <blockquote class="gmail_quote"
                                        style="margin:0 0 0 .8ex;
                                        border-left:1px #ccc solid;
                                        padding-left:1ex">
                                        <div dir="ltr">
                                          <div style="font-size:12pt;
                                            color:#000000;
                                            background-color:#ffffff;
                                            font-family:Calibri,Arial,Helvetica,sans-serif">
                                            <p>Hi,</p>
                                            <p><br>
                                            </p>
                                            <p>One of our users is
                                              running a very big
                                              simulation and writing out
                                              images of two slices (two
                                              different views) every
                                              1000 iterations and
                                              writing out the data for
                                              the two slices (two
                                              different data writers) as
                                              VTK files every 5000
                                              iterations. It is using
                                              Paraview 4.4. <br>
                                            </p>
                                            <p><br>
                                            </p>
                                            <p>After 21000 iterations,
                                              the simulation is killed
                                              because the memory on the
                                              compute nodes fills up. I
                                              usually know how to track
                                              down memory problems in
                                              our code using valgrind
                                              and related tools, but is
                                              that the right way to go
                                              to try and find this
                                              problem?</p>
                                            <p><br>
                                            </p>
                                            <p>Are there any tips on how
                                              to isolate where the
                                              problem may be? I don't
                                              know if it is in the
                                              adapter, or in paraview
                                              itself. Has anybody
                                              encountered problems with
                                              runaway memory using
                                              Catalyst in 4.4 when
                                              writing images or VTK
                                              files?
                                              <br>
                                            </p>
                                            <p><br>
                                            </p>
                                            <p>I know when we use
                                              pvpython to generate
                                              images and loop over many
                                              files, sometimes the
                                              memory also blows up and
                                              so we usually move the
                                              loop over the files
                                              outside the pvpython
                                              script and into a driver
                                              script that executes a new
                                              pvpython for each file. Is
                                              there a way to shut
                                              down/start up Catalyst
                                              each time it needs to
                                              write something? Is that
                                              advisable?
                                              <br>
                                            </p>
                                            <p><br>
                                            </p>
                                            <p>Thanks,</p>
                                            <p><br>
                                            </p>
                                            <p>Tim<br>
                                            </p>
                                          </div>
                                        </div>
                                        <br>
_______________________________________________<br>
                                        Powered by <a
                                          moz-do-not-send="true"
                                          href="http://www.kitware.com"
                                          rel="noreferrer"
                                          target="_blank"><a class="moz-txt-link-abbreviated" href="http://www.kitware.com">www.kitware.com</a></a><br>
                                        <br>
                                        Visit other Kitware open-source
                                        projects at <a
                                          moz-do-not-send="true"
                                          href="http://www.kitware.com/opensource/opensource.html"
                                          rel="noreferrer"
                                          target="_blank">
                                        </a><a moz-do-not-send="true"
                                          class="moz-txt-link-freetext"
href="http://www.kitware.com/opensource/opensource.html">http://www.kitware.com/opensource/opensource.html</a><br>
                                        <br>
                                        Please keep messages on-topic
                                        and check the ParaView Wiki at:
                                        <a moz-do-not-send="true"
                                          href="http://paraview.org/Wiki/ParaView"
                                          rel="noreferrer"
                                          target="_blank">
                                        </a><a moz-do-not-send="true"
                                          class="moz-txt-link-freetext"
href="http://paraview.org/Wiki/ParaView">http://paraview.org/Wiki/ParaView</a><br>
                                        <br>
                                        Search the list archives at: <a
                                          moz-do-not-send="true"
                                          href="http://markmail.org/search/?q=ParaView"
                                          rel="noreferrer"
                                          target="_blank">
                                        </a><a moz-do-not-send="true"
                                          class="moz-txt-link-freetext"
href="http://markmail.org/search/?q=ParaView">http://markmail.org/search/?q=ParaView</a><br>
                                        <br>
                                        Follow this link to
                                        subscribe/unsubscribe:<br>
                                        <a moz-do-not-send="true"
                                          href="http://public.kitware.com/mailman/listinfo/paraview"
                                          rel="noreferrer"
                                          target="_blank">http://public.kitware.com/mailman/listinfo/paraview</a><br>
                                        <br>
                                      </blockquote>
                                    </div>
                                    <br>
                                  </div>
                                </div>
                              </div>
                            </div>
                            <br>
                            <fieldset class="mimeAttachmentHeader"></fieldset>
                            <br>
                            <pre>_______________________________________________
Powered by <a moz-do-not-send="true" class="moz-txt-link-abbreviated" href="http://www.kitware.com">www.kitware.com</a>

Visit other Kitware open-source projects at <a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://www.kitware.com/opensource/opensource.html">http://www.kitware.com/opensource/opensource.html</a>

Please keep messages on-topic and check the ParaView Wiki at: <a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://paraview.org/Wiki/ParaView">http://paraview.org/Wiki/ParaView</a>

Search the list archives at: <a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://markmail.org/search/?q=ParaView">http://markmail.org/search/?q=ParaView</a>

Follow this link to subscribe/unsubscribe:
<a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://public.kitware.com/mailman/listinfo/paraview">http://public.kitware.com/mailman/listinfo/paraview</a>
</pre>
                          </blockquote>
                          <br>
                        </div>
                      </div>
                    </div>
                  </div>
                </div>
              </div>
            </blockquote>
            <br>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
  </body>
</html>