MPI implementation is SGI MPI. I have also tried to compile openmpi
1.4.3 and pvserver from source, yet it’s still very slow. The client
used is the x64 build from the paraview homepage.
I would expect SGI MPT to out perform others if run on an SGI system...
MPI setup is for local use only.
I'm not sure what this means?
Offscreen rendering works (I think), we have 8 windows "Visualization
Toolkit – OpenGL" that opens on the client (though the windows remain
black).
If you are running the client on one machine and the server on another,
you shouldn't see the OpenGL windows popping up on the client when you
are remote rendering. From this it sounds like all the rendering is
occurring on your client which would be very slow.
You should verify that you are not X11 forwarding, don't have -X or -Y
in your ssh commands.
You should verify that you use --use-offscreen-rendering on the pvserver
command line.
There is also one setting in the client that can cause local rendering
and you should explicitly turn it off, Edit->Settings...->Render
View->Server->Remote Render Threshold = checked, and slider is 0MB.
Also in Edit->Settings...->Render View->Server you would get improved
frame rate by using zlib when remote rendering, I'd say use at least
level 6. I usually use the presets for consumer broadband/dsl which give
a nice speed up when connecting over the internet.
Does your server system have GPUs? If not you could explicitly disable
X11 in the pvserver build. I'm pasting a configure for a remote pvserver
only build for remote offscreen only rendering no gpus.
#!/bin/bash
BUILD_TYPE=Release
INSTALL_PREFIX=/usr/common/graphics/ParaView/3.8.1
MESA_PATH=/usr/common/graphics/ParaView/Mesa-7.5.1
cmake \
-DBUILD_SHARED_LIBS=ON \
-DCMAKE_BUILD_TYPE=$BUILD_TYPE \
-DCMAKE_INSTALL_PREFIX=$INSTALL_PREFIX \
-DPARAVIEW_ENABLE_PYTHON=ON \
-DPYTHON_EXECUTABLE=/usr/common/usg/python/2.7.1/bin/python \
-DPYTHON_INCLUDE_DIR=/usr/common/usg/python/2.7.1/include/python2.7 \
-DPYTHON_LIBRARY=/usr/common/usg/python/2.7.1/lib/libpython2.7.so \
-DPARAVIEW_INSTALL_DEVELOPMENT=ON \
-DPARAVIEW_USE_MPI=ON \
-DBUILD_TESTING=OFF \
-DPARAVIEW_BUILD_QT_GUI=OFF \
-DVTK_USE_X=OFF \
-DVTK_OPENGL_HAS_OSMESA=ON \
-DOPENGL_INCLUDE_DIR=$MESA_PATH/include \
-DOPENGL_gl_LIBRARY="" \
-DOPENGL_glu_LIBRARY=$MESA_PATH/lib64/libGLU.a \
-DOPENGL_xmesa_INCLUDE_DIR=$MESA_PATH/include \
-DOSMESA_INCLUDE_DIR=$MESA_PATH/include \
-DOSMESA_LIBRARY=$MESA_PATH/lib64/libOSMesa.a \
$*
On 02/28/2011 04:15 AM, Florian Lindner wrote:
Hello,
I’m running paraview 3.8.1 under linux using 8 cpus loading a
reconstructed openfoam case with about 13GB per timestep (we have just
two time steps, start 0 and result 1). The machine on which pvserver
runs has 16 cpus and 70GB RAM, but pvserver uses just a small fraction
of the ram.
Loading the case takes about 5 minutes (which would be fine, if
subsequent operations were running more smoothly).
Adding a slice takes another couple of minutes. Rotating (using the
mouse) takes about 10sec for one frame to be rendered (the slice is
constant colored). Computing 100 streamlines from a line source takes
about 10 minutes, moving the view is fluent. As soon as the slice is
displayed moving the view becomes extremely slow again.
I have also tried to convert the mesh to VTK which does not really
change something. (Resulting VTK is of 37GB size).
At this state interactive work is not possible. Since the dataset is
not extraordinary large (I think), I suspect a problem with our MPI
configuration.
Offscreen rendering works (I think), we have 8 windows "Visualization
Toolkit – OpenGL" that opens on the client (though the windows remain
black).
MPI implementation is SGI MPI. I have also tried to compile openmpi
1.4.3 and pvserver from source, yet it’s still very slow. The client
used is the x64 build from the paraview homepage.
MPI setup is for local use only.
I am grateful for any help…
Florian
_______________________________________________
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
Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview
_______________________________________________
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
Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview
