ADVANCED_USAGE.md

Advanced usage

This page explains advanced usage scenarios for Docker images based on taltechivarlab/ubuntu-desktop.

Opening more network ports

If you intend to connect to the applications running inside this container from outside, you will likely need to open more ports than the default docker run command from our README.

💡 If you know for sure which ports in your container you will connect to, you can specify them in the initial docker run command using the -p flag. However, if you need to open the port after the container was started, please read more below.

The port mappings are specified with the initial docker run call, and you cannot delete and recreate the container without losing the data inside. Instead, please follow this answer from Stackoverflow or this article to modify the port mappings of already running container without destroying it.

Alternatively, you can run the container with the --network=host flag. This will make all ports of the container available to the host network, but only works on Linux hosts.

Enabling hardware graphics acceleration

Hardware graphics acceleration can significantly speed up your container if you are working with graphics-intensive applications or run simulations which utilize GPU for parallel computing.

As our images are based on linuxserver/rdesktop:ubuntu-mate, they support hardware graphics acceleration out of the box. To enable it, please refer to the original docs from linuxserver.io.

TL;DR:

• On any systems with Nvidia GPUs, install Nvidia container runtime on the host machine and run:

  docker run -d \
    `# all flags from the original run command here` \
    --gpus=all \
     taltechivarlab/ubuntu-desktop:20.04

  💡 You can verify that your Nvidia card was mounted successfully by running the nvidia-smi --list-gpu command in terminal inside the container. If everything went well, you will see your GPU in the list of devices output by this command. Otherwise, the command will fail.

• On Linux systems with ARM processors (e.g. Raspberry Pi):

  docker run -d \
    `# all flags from the original run command here` \
    --device=/dev/dri:/dev/dri \
     taltechivarlab/ubuntu-desktop:20.04