There are several possible ways to deploy STUnD, depending on your use case and available resources. The main supported ones are local standalone setup and containerized setup. First you have to download or clone the source code from Github, afterwards you can follow the instructions below. Both methods have been tested on Linux but should me able to be adjusted to other operating systems.
Especially for development and testing it makes most sense to run STUnD standalone locally. Before STUnD can be run it has to be compiled. The easiest way is to use the Haskell Tool Stack. Building using Cabal is not encouraged.
If you have set up Stack correctly you can simply build STUnD by running
stack build
After the build has succeeded you can simply run it using
stack run
After you see this message (Setting phasers to stun... (port 3000) (ctrl-c to quit)) you can access STUnD by directing your webbrowser to http://localhost:3000
A second way of deploying STUnD is using containers. Two common container solutions are Docker and Podman (which is mostly compatible with Docker). The latter has the advantage that the containers can be more eaily run by ordinary users. The configuration is split into two files Dockerfile and compose.yml. The first file defines how the image containing STUnD can be built. The second one describes how the container can be built on top of the image using the docker-compose or podman-compose tool.
The easiest way is just to rely on this compose tool. You can use
docker compose build # (or podman compose build)
to build the container and
docker compose up -d # (or podman compose up -d)
to run the container in the background. Afterwards STUnD can be accessed directly using a webbrowser using the URL http://localhost:3000 (port 3000 is the default and can be changed in the compose.yml file).
You can check the status of the container using
docker compose ps
and
docker compose down
can be used to stop the STUnD server.
For a more stable deployment we recommend using a reverse proxy between the user and STUnD (e.g., Apache mod_proxy or Nginx Reverse Proxy).
Docker runs containers as the root user and users have to be a member of the docker group (or similar, depending on the operating system) to be able to create and run containers. In contrast Podman runs containers as the user who created the container, which can lead to some restrictions on how containers can be configured. loginctl enable-linger.