<div dir="ltr">Takuya,<div><br></div><div> Thank you for your response. I now see how pvserver (when run in parallel) opens up the decomposed data files. </div><div><br class="">To the group,<br></div><div><br></div><div> I have sent out another email detailing the bug in native paraview openfoam reader. </div><div><br></div><div> My case is about 20 million cells and the reconstructed file sizes are 150 MB for scalar data and 500 MB for vector data. </div><div><br></div><div> I still have questions about data processing. Are filters processed on the client or server? Are complicated filters (such as stream lines that cross boundaries of the parallel (decomposed) solution) faster when pvserver is run in parallel? </div><div><br></div><div> I have re-compiled paraview 4.4.0 to be mpi-enabled. I have selected the "enable Auto MPI" with 24 processes (same number as my cores). If I open a serial (reconstructed) solution, paraview takes much longer (>10x) to open a case. Paraview is slightly slower when opening a decomposed case with the 24 processes. What should I know to improve the performance of paraview? </div><div><br></div><div>Thanks,</div><div><br></div><div>Lenny</div><div><br></div><div><br></div><div> </div></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Dec 25, 2015 at 11:45 AM, Takuya OSHIMA <span dir="ltr"><<a href="mailto:oshima@eng.niigata-u.ac.jp" target="_blank">oshima@eng.niigata-u.ac.jp</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Leonard,<br>
<br>
I am not using OpenFOAM recently so this is what I can only answer:<br>
<span class=""><br>
> Is pvserver (when<br>
> run in parallel) smart enough to not open every data file on every process for<br>
> the native openfoam reader, but to smartly allocate the directories to<br>
> different instances?<br>
<br>
</span>Yes. For the allocation strategy, see slide 7 of [1]. And if you<br>
choose "Reconstructed Case" and if the reader still works as intended,<br>
the whole recousructed case should be loaded only by process 0.<br>
<br>
[1] <a href="http://www.opencae.jp/data/OpenSourceCAEWorkshop/200811/slides/OpenSourceCAEWorkshop200811Ohshima2.pdf" rel="noreferrer" target="_blank">http://www.opencae.jp/data/OpenSourceCAEWorkshop/200811/slides/OpenSourceCAEWorkshop200811Ohshima2.pdf</a><br>
<br>
Takuya OSHIMA, Ph.D.<br>
Faculty of Engineering, Niigata University<br>
8050 Ikarashi-Ninocho, Nishi-ku, Niigata, 950-2181, JAPAN<br>
<br>
From: Leonard Cassady <<a href="mailto:lenny@intuitivemachines.com">lenny@intuitivemachines.com</a>><br>
Subject: [Paraview] Paraview 4.4 Openfoam native reader decomposed read issue<br>
Date: Wed, 23 Dec 2015 16:33:50 -0600<br>
<div class="HOEnZb"><div class="h5"><br>
> Hi, <br>
><br>
> I've recently upgraded to Openfoam 3.0 and installed paraview 4.4. To speed<br>
> up reading, I'm using pvserver in parallel (with mpirun) and open the openfoam<br>
> solution with the native openfoam reader as a "decomposed case". <br>
> I was able to visualize scalar data as expected. I was not able to render<br>
> the vector velocity data "U". It just doesn't show up as an option for<br>
> coloring surfaces or to apply filters to. Is this a bug in paraview, or am I<br>
> doing something wrong? <br>
><br>
> Above is my main question. Below are general questions regarding paraview<br>
> performance. <br>
><br>
> As you may know, Openfoam creates directories for each processor (say 192)<br>
> when solving the case in parallel. It requires the user to run the same number<br>
> of parallel processes (mpirun -np 192 <openfoamsolver> -parallel) as there<br>
> are processor directories. Does pvserver require the user to also run the<br>
> same number of processes as there are processor directories? Is pvserver (when<br>
> run in parallel) smart enough to not open every data file on every process for<br>
> the native openfoam reader, but to smartly allocate the directories to<br>
> different instances? (I know that if I open a "reconstructed case" with<br>
> pvserver running in parallel each instance of pvserver opens the same file and<br>
> greatly slows the system and overloads ram.) <br>
><br>
> I believe that my computer has good enough graphics processors to handle<br>
> rendering. Once I load in a data set I can re-orient and re-color and slice<br>
> very fast. I also believe that I can open and read the data files quickly (<br>
> about 230 MB/s read rate for my large files). I believe that the slowest part<br>
> of my data processing is the step between reading and rendering. It also<br>
> takes a long time to apply complicated filters like stream lines. How can I<br>
> speed up this part of my visualization? I have multiple computers and one of<br>
> them is much faster at analysis and I can't figure out why. <br>
> <br>
><br>
> --<br>
> Leonard Cassady PhD<br>
> Senior Development Engineer<br>
> Intuitive Machines<br>
> Cell: <a href="tel:281-755-2553" value="+12817552553">281-755-2553</a> <br>
><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature"><div dir="ltr">Leonard Cassady PhD<div>Senior Development Engineer</div><div>Intuitive Machines</div><div>Cell: 281-755-2553 </div></div></div>
</div>