Running UNIGINE-Based Application in Docker Container

You can run a UNIGINE-based application (whether it is a third-party solution or the one you've developed yourself) inside a Docker container you have already configured.

First, prepare the application that you want to run:

1. Download it from an external resource or, if it is your application, create a final Release build.

2. Move the application with all its content and additional libraries (if any) to the Docker-related folder (unigine-in-docker):

   Output

   $ mv cpp_samples/  ~/unigine-in-docker/
   $ mv csharp_component_samples/ ~/unigine-in-docker/

Running C# and C++ UNIGINE-based applications slightly differs, so here as an example we use the C++ Samples and C# Component Samples demos included in UNIGINE SDK.

1. Enable connections with the X-server:

   Output

   $ xhost +local:*

2. Launch the C++ application:

   Output

   $ cd ~/unigine-in-docker/
   $ docker run -it --rm \
   --runtime=nvidia --gpus 0 \
   -e NVIDIA_VISIBLE_DEVICES=0 \
   -e DISPLAY=${DISPLAY} \
   --device /dev/snd \
   -e NVIDIA_DRIVER_CAPABILITIES=display,compute \
   -v /tmp/.X11-unix:/tmp/.X11-unix \
   -v `pwd`/cpp_samples:/opt/project \
   -v /etc/localtime:/etc/localtime:ro \
   -w /opt/project/bin/ \
   run_unigine_in_docker:latest \
   ./cpp_samples_x64

   

1. Enable connections with the X-server:

   Output

   $ xhost +local:*

2. Launch the C# application:

   Output

   $ cd ~/unigine-in-docker/
   $ docker run -it --rm \
   --runtime=nvidia --gpus 0 \
   -e NVIDIA_VISIBLE_DEVICES=0 \
   -e DISPLAY=${DISPLAY} \
   --device /dev/snd \
   -e NVIDIA_DRIVER_CAPABILITIES=display,compute \
   -v /tmp/.X11-unix:/tmp/.X11-unix \
   -v `pwd`/csharp_component_samples:/opt/project \
   -v /etc/localtime:/etc/localtime:ro \
   -w /opt/project/bin/ \
   run_unigine_in_docker:latest \
   dotnet csharp_component_samples_x64.dll -console_command "world_load csharp_samples"

   

Congratulations! You have successfully run a UNIGINE-based application in a Docker container!