<html>
  <head>
    <meta content="text/html; charset=windows-1252"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <tt>Joshua -<br>
      <br>
      It does not seem you are attempting to use large blocks, but the
      problem likely is computing block geometry in the case the mesh is
      deeply nested. 2^30 factor is pretty big and a loss of accuracy is
      nearly inevitable in this case since Paramesh operates on real,
      physical coordinates. Remove a few refinement levels and see if
      the problem goes away.<br>
      <br>
      So my feeling is that this is not the interpolation package
      itself. :-)<br>
      <br>
      Tomek<br>
      --</tt><br>
    <div class="moz-cite-prefix">On 04/23/15 10:54, Joshua Wall wrote:<br>
    </div>
    <blockquote
cite="mid:CAHnUEowR-Ezouhcakd4EXg3De+B+pjs+hspVU--07xQM8sRa2Q@mail.gmail.com"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=windows-1252">
      <div dir="ltr">
        <div>Tomasz,<br>
          <br>
        </div>
            My setup line is:<br>
        <br>
        ./setup SinkRotating -a -3d -site=draco
        -parfile="flash.par.1000M" +uhd<br>
        <div>
          <div><br>
          </div>
          <div>where I've modified the par file to make the cloud 1000M
            and extended the refinement to 30 levels (along with the
            required changes to the sink threshold density and accr
            radius).<br>
            <br>
          </div>
          <div>Josh<br>
          </div>
          <div><br>
            <div class="gmail_quote">On Thu, Apr 23, 2015 at 10:18 AM
              Tomasz Plewa <<a moz-do-not-send="true"
                href="mailto:tplewa@fsu.edu">tplewa@fsu.edu</a>>
              wrote:<br>
              <blockquote class="gmail_quote" style="margin:0 0 0
                .8ex;border-left:1px #ccc solid;padding-left:1ex">
                <div bgcolor="#FFFFFF" text="#000000"> <tt>Dear Joshua
                    - <br>
                    <br>
                    Multidimensional conservative mapping is
                    complicated. The original package came from NCAR,
                    and adapting and extending it to the needs of hydro
                    applications (first AMRA, then FLASH) was quite an
                    adventure.<br>
                  </tt><br>
                  I have seen this kind of error ages ago. It means that
                  coordinates are non-monotonic.  The most likely cause
                  is <tt><tt>the geometry of your blocks is not passed
                      correctly to the mapping package.<br>
                      <br>
                    </tt>What is your setup line? Some memory-related
                    entities are hardwired and you might simply be
                    exceeding those limits.<br>
                    <br>
                    Tomek<br>
                    --</tt></div>
                <div bgcolor="#FFFFFF" text="#000000"><br>
                  <div>On 04/23/15 09:14, Joshua Wall wrote:<br>
                  </div>
                  <blockquote type="cite">
                    <div dir="ltr">
                      <div>
                        <div>
                          <div>I've been tracking the error down a bit
                            more, here's what I can add.<br>
                            <br>
                          </div>
                              The call for Driver_abortFlash is coming
                          from amr_prolong_gen_unk1_fun.F90 which is
                          trying to call the umap3 subroutine in umap.F
                          (for interpolation between different sized
                          grids?). Error 421 specifically states in
                          umap3:<br>
                          <br>
                                  ier = 421<br>
                                  do ii = 2,mx<br>
                                     if ( xx(ii-1).gt.xx(ii) ) return<br>
                                  end do<br>
                          <br>
                        </div>
                        There are a LOT of variables in this subroutine,
                        and its a bit hard to follow what this is
                        checking. So I'm having trouble parsing out much
                        past this point. I might add that these runs are
                        across multiple nodes on a Beowulf cluster (the
                        first I've attempted like this) and its the
                        first time I've seen this error. Oddly enough,
                        my first run was fine until I got a hang during
                        a refinement iteration (probably because I
                        either ran out of processors or memory). Then
                        when I tried to make a few changes to add some
                        procs, I have gotten this error. Any help would
                        be greatly appreciated.<br>
                        <br>
                      </div>
                      Josh<br>
                    </div>
                    <br>
                    <div class="gmail_quote">On Wed, Apr 22, 2015 at
                      6:53 PM Joshua Wall <<a moz-do-not-send="true"
                        href="mailto:joshua.e.wall@gmail.com"
                        target="_blank">joshua.e.wall@gmail.com</a>>
                      wrote:<br>
                      <blockquote class="gmail_quote" style="margin:0 0
                        0 .8ex;border-left:1px #ccc
                        solid;padding-left:1ex">
                        <div dir="ltr">
                          <div>
                            <div>
                              <div>Dear Users,<br>
                                <br>
                              </div>
                                  I'm receiving an odd error from FLASH
                              and I'm not sure how to interpret it.
                              Perhaps someone can enlighten me? Its:<br>
                              <br>
                               **ERROR** - ierr from UMAPn is         
                              421 , isg=           4<br>
                               DRIVER_ABORT: ERROR - ierr from UMAPn is
                              not 0<br>
                              <br>
                            </div>
                            Cordially,<br>
                            <br>
                          </div>
                          Josh<br>
                        </div>
                      </blockquote>
                    </div>
                  </blockquote>
                  <br>
                </div>
              </blockquote>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
  </body>
</html>