<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Aug 15, 2016 at 9:28 PM, leejearl <span dir="ltr"><<a href="mailto:leejearl@126.com" target="_blank">leejearl@126.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
  
    
  
  <div bgcolor="#FFFFFF" text="#000000">
    <p>Thank you for all your helps. I have tried and reinstalled PETSc
      lots of times. The error are still existed.</p>
    <p>I can not find out the reasons. Now, I give the some messages in
      this letter.</p>
    <p>1> The source code is downloaded from the website of PETSc,
      and the version is 3.7.2.<br>
    </p>
    <p></p></div></blockquote><div>You will need to run in the 'master' branch, not the release since we have fixed some bugs. It is best to</div><div>use master for very new features like this. The instructions are here: <a href="http://www.mcs.anl.gov/petsc/developers/index.html">http://www.mcs.anl.gov/petsc/developers/index.html</a></div><div><br></div><div>   Thanks,</div><div><br></div><div>      Matt <br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div bgcolor="#FFFFFF" text="#000000"><p>2> Configure: <br>
    </p>
    <p>>export PETSC_DIR=./</p>
    <p>>export PETSC_ARCH=arch<br>
    </p>
    <p>>./configure --prefix=$HOME/Install/petsc
      --with-mpi-dir=/home/leejearl/<wbr>Install/mpich_3.1.4/gnu
      --download-exodusii=../<wbr>externalpackages/exodus-5.24.<wbr>tar.bz2
      --download-netcdf=../<wbr>externalpackages/netcdf-4.3.2.<wbr>tar.gz
      --download-hdf5=../<wbr>externalpackages/hdf5-1.8.12.<wbr>tar.gz
      --download-metis=../<wbr>externalpackages/git.metis.<wbr>tar.gz
      --download-parmetis=yes<br>
    </p>
    <p>3> The process of installation has no error.</p>
    <p>4> After the installation, I added the following statement
      into the file ~/.bashrc:</p>
    <p>export PETSC_ARCH=""</p>
    <p>export PETSC_DIR=$HOME/Install/pets/ <br>
    </p>
    <p>I wish to get some helps as follows:</p>
    <p>1> Is there any problems in my installation? <br>
    </p>
    <p>2> Can any one help me a simple code in which the value of
      overlap used in DMPlexDistribute function is greater than 1.</p>
    <p>3> I attach the code, makefile, grid and the error messages
      again, I hope some one can help me to figure out the problems.</p>
    <p>    3.1> code: cavity.c</p>
    <p>    3.2> makefile: makefile</p>
    <p>    3.3> grid: cavity.exo</p>
    <p>    3.4> error messages: error.dat</p>
    <p>    It is very strange that there is no error message when I run
      it using "mpirun -n 3 ./cavity", but when I run it using "mpirun
      -n 2 ./cavity", the errors happed.</p>
    <p>The error messages are shown in the file error.dat.<br>
    </p>
    <p><br>
    </p>
    <p>Any helps are appreciated.</p>
    <p><br>
    </p>
    <br>
    <div>On 2016年08月13日 09:04, leejearl wrote:<br>
    </div>
    <blockquote type="cite">
      
      <p>Thank you for your reply. The source code I have used is from
        the website of PETSc, not from the git repository.</p>
      <p>I will have a test with the code from git repository.<br>
      </p>
      <p><br>
      </p>
      <p>leejearl<br>
      </p>
      <br>
      <div>On 2016年08月13日 08:49, Oxberry,
        Geoffrey Malcolm wrote:<br>
      </div>
      <blockquote type="cite">
        
        <br>
        <div>
          <blockquote type="cite">
            <div>On Aug 12, 2016, at 5:41 PM, leejearl <<a href="mailto:leejearl@126.com" target="_blank">leejearl@126.com</a>> wrote:</div>
            <br>
            <div>
              <div bgcolor="#FFFFFF" text="#000000">
                <p>Hi, Matt:<br>
                </p>
                    <br>
                <div><br>
                </div>
                > Can you verify that you are running the master
                branch?<br>
              </div>
            </div>
          </blockquote>
          <div><br>
          </div>
          <div>cd ${PETSC_DIR}</div>
          <div>git branch</div>
          <div><br>
          </div>
          <div>The last command should return something like a list of
            branch names, and the branch name with an asterisk to the
            left of it will be the branch you are currently on.</div>
          <div><br>
          </div>
          <div>Geoff</div>
          <div>
            <div><br>
            </div>
          </div>
          <blockquote type="cite">
            <div>
              <div bgcolor="#FFFFFF" text="#000000">I am not
                sure, how can I verify this?<br>
                And I configure PETSc with this command <br>
                "./configure --prefix=$HOME/Install/petsc-<wbr>openmpi
                --with-mpi=/home/leejearl/<wbr>Install/openmpi/gnu/1.8.4
                --download-exodusii=yes --download-netcdf
                --with-hdf5-dir=/home/<wbr>leejearl/Install/hdf5-1.8.14
                --download-metis=yes".<br>
                Is there some problem? Can you show me your command for
                configuring PETSc?<br>
                <br>
                <br>
                Thanks<br>
                <br>
                leejearl<br>
                <br>
                <br>
                <br>
                <br>
                <br>
                <div>On 2016年08月13日 01:10,
                  Matthew Knepley wrote:<br>
                </div>
                <blockquote type="cite">
                  <div dir="ltr">
                    <div class="gmail_extra">
                      <div class="gmail_quote">On Thu, Aug 11, 2016 at
                        8:00 PM, leejearl <span dir="ltr">
                          <<a href="mailto:leejearl@126.com" target="_blank">leejearl@126.com</a>></span>
                        wrote:<br>
                        <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                          <div bgcolor="#FFFFFF" text="#000000">
                            <p>Thank you for your reply. I have
                              attached the code, grid and the error
                              message.</p>
                            <p>cavity.c is the code file,
                              cavity.exo is the grid, and error.dat is
                              the error message.</p>
                            <p>The command is "mpirun -n 2
                              ./cavity</p>
                          </div>
                        </blockquote>
                        <div><br>
                        </div>
                        <div>Can you verify that you are
                          running the master branch? I just ran this and
                          got</div>
                        <div><br>
                        </div>
                        <div>
                          <div>DM Object: 2 MPI processes</div>
                          <div>  type: plex</div>
                          <div>DM_0x84000004_0 in 2 dimensions:</div>
                          <div>  0-cells: 5253 5252</div>
                          <div>  1-cells: 10352 10350</div>
                          <div>  2-cells: 5298 (198) 5297 (198)</div>
                          <div>Labels:</div>
                          <div>  ghost: 2 strata of sizes (199,
                            400)</div>
                          <div>  vtk: 1 strata of sizes (4901)</div>
                          <div>  Cell Sets: 1 strata of sizes
                            (5100)</div>
                          <div>  Face Sets: 3 strata of sizes
                            (53, 99, 50)</div>
                          <div>  depth: 3 strata of sizes
                            (5253, 10352, 5298)</div>
                        </div>
                        <div><br>
                        </div>
                        <div>  Thanks,</div>
                        <div><br>
                        </div>
                        <div>     Matt</div>
                        <div> </div>
                        <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                          <div bgcolor="#FFFFFF" text="#000000">
                            <div>On 2016年08月11日 23:29, Matthew
                              Knepley wrote:<br>
                            </div>
                            <blockquote type="cite">
                              <div dir="ltr">
                                <div class="gmail_extra">
                                  <div class="gmail_quote">On Thu, Aug
                                    11, 2016 at 3:14 AM, leejearl <span dir="ltr"> <<a href="mailto:leejearl@126.com" target="_blank">leejearl@126.com</a>></span>
                                    wrote:<br>
                                    <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                                      <div bgcolor="#FFFFFF" text="#000000">
                                        <p>Hi, <br>
                                              Thank you for your reply.
                                          It help me very much.<br>
                                              But, for
                                          "/petsc-3.7.2/src/ts/examples/<wbr>tutorials/ex11.c",
                                          when I set the overlap to 2
                                          levels, the command is <br>
                                          "mpirun -n 3 ./ex11 -f
                                          annulus-20.exo
                                          -ufv_mesh_overlap 2 -physics
                                          sw", it suffers a error.<br>
                                              It seems to me that
                                          setting overlap to 2 is very
                                          common. Are there issues that
                                          I have not take into
                                          consideration?<br>
                                              Any help are appreciated.</p>
                                      </div>
                                    </blockquote>
                                    <div>I will check this out.
                                      I have not tested an overlap of 2
                                      here since I generally use nearest
                                      neighbor FV methods for</div>
                                    <div>unstructured stuff. I
                                      have test examples that run fine
                                      for overlap > 1. Can you send
                                      the entire error message?</div>
                                    <div><br>
                                    </div>
                                    <div>If the error is not in
                                      the distribution, but rather in
                                      the analytics, that is
                                      understandable because this
                                      example is only</div>
                                    <div>intended to be run
                                      using a nearest neighbor FV
                                      method, and thus might be confused
                                      if we give it two layers of ghost</div>
                                    <div>cells.</div>
                                    <div><br>
                                    </div>
                                    <div>   Matt</div>
                                    <div> </div>
                                    <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                                      <div bgcolor="#FFFFFF" text="#000000">
                                        <p><span><font color="#888888"><br>
                                              leejearl</font></span></p>
                                        <div>
                                          <div><br>
                                            <div>On 2016年08月11日
                                              14:57, Julian Andrej
                                              wrote:<br>
                                            </div>
                                            <blockquote type="cite">
                                              <div dir="ltr">Hi,
                                                <div><br>
                                                </div>
                                                <div>take a
                                                  look at slide 10 of
                                                  [1], there is visually
                                                  explained what the
                                                  overlap between
                                                  partitions is.</div>
                                                <div><br>
                                                </div>
                                                <div>[1] <a href="https://www.archer.ac.uk/training/virtual/files/2015/06-PETSc/slides.pdf" target="_blank">https://www.archer.ac.uk/t<wbr>raining/virtual/files/2015/06-<wbr>PETSc/slides.pdf</a></div>
                                              </div>
                                              <div class="gmail_extra"><br>
                                                <div class="gmail_quote">On
                                                  Thu, Aug 11, 2016 at
                                                  8:48 AM, leejearl <span dir="ltr">
                                                    <<a href="mailto:leejearl@126.com" target="_blank">leejearl@126.com</a>></span>
                                                  wrote:<br>
                                                  <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                                                    Hi, all:<br>
                                                        I want to use
                                                    PETSc to build my
                                                    FVM code. Now, I
                                                    have a question
                                                    about<br>
                                                    the function 
                                                    DMPlexDistribute(DM
                                                    dm, PetscInt
                                                    overlap, PetscSF
                                                    *sf, DM *dmOverlap)
                                                    .<br>
                                                    <br>
                                                        In the example
                                                    "/petsc-3.7.2/src/ts/examples/<wbr>tutorials/ex11.c".
                                                    When I set the
                                                    overlap<br>
                                                    as 0 or 1, it works
                                                    well. But, if I set
                                                    the overlap as 2, it
                                                    suffers a problem.<br>
                                                        I am confused
                                                    about the value of
                                                    overlap. Can it be
                                                    set as 2? What is
                                                    the meaning of<br>
                                                    the parameter
                                                    overlap?<br>
                                                        Any helps are
                                                    appreciated!<span><font color="#888888"><br>
                                                        <br>
                                                        leejearl<br>
                                                        <br>
                                                        <br>
                                                        <br>
                                                      </font></span></blockquote>
                                                </div>
                                                <br>
                                              </div>
                                            </blockquote>
                                            <br>
                                          </div>
                                        </div>
                                      </div>
                                    </blockquote>
                                  </div>
                                  <br>
                                  <br clear="all"><span class=""><font color="#888888">
                                  <span><font color="#888888">
                                      <div><br>
                                      </div>
                                      -- <br>
                                      <div data-smartmail="gmail_signature">What most experimenters
                                        take for granted before they
                                        begin their experiments is
                                        infinitely more interesting than
                                        any results to which their
                                        experiments lead.<br>
                                        -- Norbert Wiener</div>
                                    </font></span></font></span></div><span class=""><font color="#888888">
                              </font></span></div><span class=""><font color="#888888">
                            </font></span></blockquote><span class=""><font color="#888888">
                            <br>
                          </font></span></div><span class=""><font color="#888888">
                        </font></span></blockquote><span class=""><font color="#888888">
                      </font></span></div><span class=""><font color="#888888">
                      <br>
                      <br clear="all">
                      <div><br>
                      </div>
                      -- <br>
                      <div data-smartmail="gmail_signature">What most
                        experimenters take for granted before they begin
                        their experiments is infinitely more interesting
                        than any results to which their experiments
                        lead.<br>
                        -- Norbert Wiener</div>
                    </font></span></div><span class=""><font color="#888888">
                  </font></span></div><span class=""><font color="#888888">
                </font></span></blockquote><span class=""><font color="#888888">
                <br>
              </font></span></div><span class=""><font color="#888888">
            </font></span></div><span class=""><font color="#888888">
          </font></span></blockquote><span class=""><font color="#888888">
        </font></span></div><span class=""><font color="#888888">
        <br>
      </font></span></blockquote>
    </blockquote>
    <br>
  </div>

</blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature" data-smartmail="gmail_signature">What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>-- Norbert Wiener</div>
</div></div>