Skip to content

Running simulations on a remote server

Jump to bottom
Buridi Aditya edited this page Oct 9, 2018 · 1 revision

Setup in Linux remote server with GPU:

** WILL NEED SUPERUSER PERMISSIONS **

  1. Install VirtualGL and TurboVNC. (Make sure you install the 64bit version of virtualgl).
  2. Shut down the display manager: /etc/init.d/gdm stop or /etc/init.d/lightdm stop depending on the display manager used in your linux system (how to figure this out? Only one of the service will be available try ls /etc/init.d/ to find out the right one ).
  3. Log in as root from the text console (or remotely using ssh.)
  4. Run /opt/VirtualGL/bin/vglserver_config
  5. Select option 1 (Configure the server for use with VirtualGL in GLX mode.)
  6. Restrict 3D X server access to vglusers group (recommended)? [Y/n] Enter Y.
  7. Restrict framebuffer device access to vglusers group (recommended)? [Y/n] Enter Y.
  8. Disable XTEST extension (recommended)? [Y/n] Enter N.
  9. Open /etc/group and search for vglusers and append root,username in that line.
  10. Restart the display manager /etc/init.d/gdm start or /etc/init.d/lightdm start

On Linux client machine:

  1. Install VirtualGL andTurboVNC. (Make sure you install the 64bit version of virtualgl).
  2. For running an OpenGL specific application in the server run vglconnect -s username@<remote-server>
  3. Run vglrun <application>.

In case you end up with the error [VGL] ERROR: Could not open display :0. run sudo /etc/init.d/lightdm restart or sudo /etc/init.d/gdm restart.

References:

  1. https://summerofhpc.prace-ri.eu/remote-accelerated-graphics-with-virtualgl-and-turbovnc/
  2. https://virtualgl.org/vgldoc/2_2_1/
Clone this wiki locally