Creating A New Docker Image

Creating a Docker Image

OpenBoxes release images are automatically published to GitHub packages whenever a commit is tagged for release (see the workflow file for details). No manual work required.

However, if you ever need to manually create a Docker image for the application, you can do so by following these instructions:

Prerequisites

1. Install Docker (20.04+)

2. Clone the openboxes repository

3. Set up an openboxes database with a valid db user

Steps

1. Start Docker

2. Navigate to the root directory of the openboxes repository

3. Run ./gradlew prepareDocker -Dgrails.env=prod

   1. This produces an executable openboxes.war file, then copies it and its configuration files to the /build/docker directory. Note that the generated WAR file contains an embedded Tomcat servlet.

4. Run docker build --tag="openboxes/openboxes:latest" build/docker/

   1. This will take in the WAR file and build a Docker image for the app.

Running Docker Images

We provide Docker Compose files that greatly simplify the process of starting up your application containers and hooking them up to your SQL server. We recommend using the compose files instead of trying to manually run and configure the containers yourself.

Once you have a Docker image (whether it's an official release or one you created yourself) you can use it to start up a containerized version of OpenBoxes.

We assume that you have already configured your database server by this point.

If running an official release image, run the following:

• docker run -p 8080:8080 --name=openboxes ghcr.io/openboxes/openboxes:latest

  • If you need a specific version, you can replace latest with that version (such as v0.9.5)

If running a manually created image, run the following:

• docker run -p 8080:8080 --name=openboxes openboxes/openboxes:latest

Then navigate to http://localhost:8080/openboxes to access your server.

Overriding Environment Variables

When running the application locally, the default environment variables are likely adequate, but if you need to customize them for your specific setup, you can do so via a .env file.

• Create a docker/.env file (see the .env.example file for reference)

• Add the --env-file parameter to your docker run command

  • For example: docker run --env-file=docker/.env -p 8080:8080 --name=openboxes openboxes/openboxes:latest

Alternatively, you can directly provide environment variables via the --env parameter.

For example: docker run -p 8080:8080 --name=openboxes --env DATASOURCE_USERNAME='openboxes' --env=DATASOURCE_PASSWORD='openboxes' openboxes/openboxes:latest

Persistable Logs (journald)

If you'd like the application logs to be persisted after the container has been stopped/removed, run the docker container with journald agent.

For example: docker run --env-file=docker/.env -p 8080:8080 --name=openboxes --log-driver=journald openboxes/openboxes:latest

The logs will then be able to be accessed with: journalctl -u docker CONTAINER_NAME=openboxes

Running against a database server on the host machine

This process is much simpler via use of the Docker Compose files, and is specifically handled for you in docker-compose-hostdb.yml

You can connect a containerized OpenBoxes instance to a SQL server that is running on the host by using the host.docker.internal keyword in the JDBC url.

To achieve this, override the DATASOURCE_URL environment variable and add the host gateway via the --add-host parameter.

For example: docker run -p 8080:8080 --name=openboxes --env DATASOURCE_URL='jdbc:mysql://host.docker.internal:3306/openboxes?useSSL=false' --add-host host.docker.internal:host-gateway openboxes/openboxes:latest

You will also need to set bind.address=0.0.0.0 in your SQL configuration.