[Paraview-developers] pvserver with multiple EGL devices

[Dan Lipsa]

Jean,
If you use only
mpirun  -np 1 ./bin/pvserver --egl-device-index=1. Does that work
corrrectly?
Can you see in client that it uses device 1?
Help, About, Connection Information, OpenGL Renderer.
I know we had a bug with how this parameter was passed, but the bug has
been fixed a while ago.


What about
mpirun  -np 1 ./bin/pvserver
(which should use device 0). Does that work?

Thanks,
Dan



On Wed, Oct 11, 2017 at 11:03 AM, Favre Jean <jfavre at cscs.ch> wrote:

> Folks
>
> I've used EGL for a long time now, but always with a single GPU. I
> understand it well.
>
> I have a new cluster with *two* GPUs on the node and a more recent driver,
> where I compiled 5.4 with EGL and OpenMPI
>
> Launching pvserver, and connecting I get this error:
>
> mpirun  -np 1 ./bin/pvserver --egl-device-index=0  : -np 1 ./bin/pvserver
> --egl-device-index=1
>
> Waiting for client...
> Connection URL: cs://greina28:11111
> Accepting connection(s): greina28:11111
> Client connected.
> Warning: In /users/jfavre/ParaView/ParaView-v5.4.1/VTK/Rendering/
> OpenGL2/vtkEGLRenderWindow.cxx, line 272
> vtkEGLRenderWindow (0xed65a0): Setting an EGL display to device index: 0
> require EGL_EXT_device_base EGL_EXT_platform_device EGL_EXT_platform_base
> extensions
>
> ERROR: In /users/jfavre/ParaView/ParaView-v5.4.1/VTK/Rendering/
> OpenGL2/vtkEGLRenderWindow.cxx, line 341
> vtkEGLRenderWindow (0xed65a0): Only EGL 1.4 and greater allows OpenGL as
> client API. See eglBindAPI for more information.
>
> ERROR: In /users/jfavre/ParaView/ParaView-v5.4.1/VTK/Rendering/
> OpenGL2/vtkOpenGLRenderWindow.cxx, line 819
> vtkEGLRenderWindow (0xed65a0): GLEW could not be initialized.
>
> yet, my ./bin/pvserver depends on
>
>     libEGL.so.1 => /usr/lib64/libEGL.so.1 (0x00002aaabc2c8000)
>     libOpenGL.so.0 => /usr/lib64/libOpenGL.so.0 (0x00002aaabc4f5000)
>     libGLdispatch.so.0 => /usr/lib64/libGLdispatch.so.0
> (0x00002aaabc72a000)
>
> and nvidia-smi reports a driver which should support EGL1.4 and greater.
>
> Wed Oct 11 17:00:24 2017
> +-----------------------------------------------------------
> ------------------+
> | NVIDIA-SMI 384.59                 Driver Version:
> 384.59                    |
> |-------------------------------+----------------------+----
> ------------------+
> | GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr.
> ECC |
> | Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute
> M. |
> |===============================+======================+====
> ==================|
> |   0  Tesla P100-SXM2...  On   | 00000000:07:00.0 Off |
> Off |
> | N/A   30C    P0    30W / 300W |      0MiB / 16276MiB |      0%
> Default |
> +-------------------------------+----------------------+----
> ------------------+
> |   1  Tesla P100-SXM2...  On   | 00000000:84:00.0 Off |
> Off |
> | N/A   32C    P0    31W / 300W |      0MiB / 16276MiB |      0%
> Default |
> +-------------------------------+----------------------+----
> ------------------+
>
>
> what setting should I be specifically careful with to overcome this error?
>
> TIA
> -----------------
> Jean / CSCS
>
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/
> opensource/opensource.html
>
> Search the list archives at: http://markmail.org/search/?q=
> Paraview-developers
>
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/paraview-developers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview-developers/attachments/20171011/7ddbce81/attachment-0001.html>