[Paraview] Correctly Parallel Processing of OpenFoam results using pvserver

Ganesh Vijayakumar ganesh.iitm at gmail.com
Fri Oct 2 10:22:09 EDT 2015


On Wed, Sep 30, 2015 at 1:35 PM, Leonard Cassady <
lenny at intuitivemachines.com> wrote:

> I'm attempting to use pvserver to accelerate the post-processing of my
> openfoam solution. I have a 48 core machine. I have correctly installed and
> compiled a parallel copy of paraview 4.1.0 with OpenFOAM 2.4.x. If I open a
> simple .obj file I can see that different parts of the surface are rendered
> using different processors. I can also see that the memory is shared among
> the parallel processes.
>
> When I open a reconstructed openFOAM solution with 20 million cells with
> paraview connected to 40 process pvserver, the image seems to be rendered
> (or processed) with only 1 processor. Is there a step that I'm missing to
> parallelize the reconstructed Openfoam data files for rendering?
>

See if you have to reconstruct it. You may be able to get much faster
visualization with your setup (I/O not withstanding) if you just visualize
the decomposed case. Choosing the "case type" to "decomposed" will do it.

 Inorder to check whether one or multiple processes are actually handing
the data, make a contour plot of the node ID or some variable like that.
This should most likely not be a part of your actual OpenFOAM solution.
Paraview seems to assign the processor ID to each cell.
Plotting/Visualizing that may be a good test of whether you're actually
using Paraview's parallel capabilities or not.

ganesh


>
> --
> Leonard Cassady PhD
> Senior Development Engineer
> Intuitive Machines
> Cell: 281-755-2553
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ParaView Wiki at:
> http://paraview.org/Wiki/ParaView
>
> Search the list archives at: http://markmail.org/search/?q=ParaView
>
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview
>
>


-- 
ganesh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20151002/7f6ac0f2/attachment.html>


More information about the ParaView mailing list