This Code Defenders, a mutation testing game.
A public instance is available at code-defenders.org and the source code can be found at GitHub.

We provide a docker-compose file for a simple setup.
Documentation for getting started can be found below.

We (more or less regularly) publish images to:

• quay.io/codedefenders/codedefenders
• docker.io/codedefenders/codedefenders
• ghcr.io/codedefendes/codedefenders

We publish images for tagged Code Defenders versions/releases (starting at version 1.8).
But also sometimes (git-<short ref hash>) tags, build of the corresponding git commit.

If you want to build images locally take a look at the Building Container Images section.

The docker-compose file is available in our GitHub repository at docker/docker-compose.yml.
The easiest way to get it is to check out the repository locally.

Otherwise, download the docker-compose.yml and example.env file, like:

mkdir docker
curl https://raw.githubusercontent.com/CodeDefenders/CodeDefenders/master/docker/docker-compose.yml > docker/docker-compose.yml
curl https://raw.githubusercontent.com/CodeDefenders/CodeDefenders/master/docker/example.env > docker/example.env

Copy the docker/example.env file to docker/.env and update the CODEDEFENDERS_VERSION variable to the tag of the image you want to use.

Now you can

cd docker
docker-compose up

to spin up the containers.
By default, Code Defenders will be available at http://localhost:8080/ in your browser.

Note: It is a good idea to check out the corresponding git tag (in this example v1.8.1), as it is possible that the docker-compose setup from two versions ago does not work with the docker image from the current version.

Configuring the docker-compose setup is possible via environment variables.
You can find all available docker compose specific variables with their default values in the example.env file.

For environment variables which are available by default see the configuration documentation.

The docker container supports only a subset of the available configuration properties.
E.g. all cluster.* related properties (inclusive force.local.execution) are not supported because we have no slurm support inside the container.

WARNING:
Changing the CODEDEFENDERS_DB_* variables after creating the container for the first time requires you to update the settings in the database container manually.
Otherwise, you have to destroy the volumes.

NOTE:
The container supports the intentionally undocumented CODEDEFENDERS_TOMCAT_LISTENING_PORT variable to change the port the tomcat server in the container binds to.
It can be used to prevent port collisions for setups where the codedefenders containers are run with the (less secure) --network=host option.

The docker-compose file uses named volumes to persist the database and codedefenders data directory.
The names used in a normal setup should be docker_datavolume and docker_dbvolume.

Those are managed via the docker volume command.

As long as those volumes are there, the content of the mysql db and the data folder are persisted. So if you have stopped, and even destroyed, the containers by using the same docker-compose file the data are preserved.

WARNING:
You should only ever rm both volumes at the same time as otherwise the database will reference no longer existing files, or the database will try to write at places where already some files exist.

More information can be found in the docker documentation

A minimum of reliability is guaranteed by automatically restarting the container if the fail or are manually stopped (by mistake). Therefore, unless you stop the docker-compose, containers will automatically restart.

Note Restarting the containers might not preserve the user sessions, that is, your users might have to re-login in the system after a crash/restart.

If you need more instances or if you need to customize the deployment, you must update the Docker and various configurations files under the /docker folder. But at that point you are on your own.

We build from the git repository root, so we can simply copy the source code for building.

cd <path to codedefenders repository>
docker build --file ./docker/Dockerfile --tag codedefenders/codedefenders:<Codedefenders version> .

The official images are build (and published) from the CI with the help of the build-container-image and push-container-images scripts (Repository local links).

You can tag a locally built docker image and use this for the docker-compose.
Be aware that we use named docker volumes which are persisted on the host.
So building new images and deploying them will still get the old data.

cd <path to codedefenders repository>
docker build --file ./docker/Dockerfile --tag localhost/codedefenders:dev .
cd docker
nano .env # Set CODEDEFENDERS_VERSION=dev and CODEDEFENDERS_IMAGE=localhost/codedefenders
docker-compose up