Review Bot Docker Images¶

New in version 3.0.

Review Bot can be installed through our official Docker images, helping you quickly get your infrastructure set up with minimal effort.

Each release ships with a set of pre-made tool pack images that check certain languages. You can also create your own custom Review Bot images that contain just the tools you need. This is all covered below.

Getting Started¶

We’ll walk you through the basics of setting up Review Bot on Docker in two different ways:

Launching the Review Bot image itself using docker run.
Launching the image along with a database, web server, and memcached using docker-compose.

You’ll also see how you can build a Docker image for just the tools you need.

Before You Begin¶

Plan Your Infrastructure¶

Review Bot requires a compatible message broker.

RabbitMQ is most commonly used, and is available through the rabbitmq Docker image. This is what we’ll use for this guide.

Choose a Version and Tool Packs¶

Our Docker images are available in the following forms:

beanbag/reviewbot-pack:latest – The latest stable release of Review Bot for a given language/tool pack.
beanbag/reviewbot-pack:X – The latest stable release in a major version series (e.g., 3).
beanbag/reviewbot-pack:X.Y – The latest stable release in a major.minor version series (e.g., 3.0).
beanbag/reviewbot-pack:X.Y.Z – A specific release of Review Bot (e.g., 3.0.1).

Note

We recommend using a {X.Y}-versioned image, rather than using latest, to help avoid unexpected changes with future versions. You’ll still benefit from bug fixes.

You can choose any of the following tool pack images:

Using Docker Run¶

First, make sure you have RabbitMQ (or another compatible message broker) configured and ready for Review Bot. You can use an existing Docker image or install a broker manually.

To start a new container, run:

$ docker pull beanbag/reviewbot-<pack>:X.Y.Z
$ docker run -P \
      --name <name> \
      -v <local_path>:/config \
      -v <local_path>:/repos \
      -e BROKER_URL=<broker URL> \
      beanbag/reviewbot-<pack>:X.Y.Z

For example:

$ docker pull beanbag/reviewbot-python:3.0
$ docker run -P \
      --name <name> \
      -v /etc/reviewbot/config:/config \
      -v /var/lib/reviewbot/repos:/repos \
      -e BROKER_URL=amqp://reviewbot:reviewbot123@rabbitmq/reviewbot \
      beanbag/reviewbot-python:3.0

We’ll go over the settings later.

Using Docker Compose¶

docker-compose can help you define and launch all the services needed for your Review Bot deployment.

A simple docker-compose.yaml for launching RabbitMQ and two Review Bot workers (one for Python, one for Go) might look like:

docker-compose.yaml¶

version: '3.7'

services:
  rabbitmq:
    image: rabbitmq:3-management
    restart: always
    hostname: rabbitmq
    environment:
      - RABBITMQ_DEFAULT_VHOST=reviewbot
      - RABBITMQ_DEFAULT_USER=reviewbot
      - RABBITMQ_DEFAULT_PASS=reviewbot123
      - RABBITMQ_ERLANG_COOKIE=secret-dont-tell
    volumes:
      - /var/lib/reviewbot/rabbitmq/data:/var/lib/rabbitmq/
      - /var/log/reviewbot/rabbitmq:/var/log/rabbitmq
    ports:
      - 15672:15672
      - 5672:5672
    healthcheck:
      test: ['CMD', 'rabbitmqctl', 'status']
      interval: 5s
      timeout: 20s
      retries: 5

  reviewbot-go:
    image: beanbag/reviewbot-go:3.0
    restart: always
    depends_on:
      rabbitmq:
        condition: service_healthy
    environment:
      - BROKER_URL=amqp://reviewbot:reviewbot123@rabbitmq/reviewbot
    volumes:
      - /etc/reviewbot/config:/config
      - /var/lib/reviewbot/repos:/repos


  reviewbot-python:
    image: beanbag/reviewbot-python:3.0
    restart: always
    depends_on:
      rabbitmq:
        condition: service_healthy
    environment:
      - BROKER_URL=amqp://reviewbot:reviewbot123@rabbitmq/reviewbot
    volumes:
      - /etc/reviewbot/config:/config
      - /var/lib/reviewbot/repos:/repos

You’ll want to tailor this to your configuration and security requirements. This is purely for demonstrative purposes.

To bring up the environment, run:

$ docker-compose up

You can then enable Review Bot in Review Board and point it to the broker URL for your server.

Note

Make sure to use the publicly-resolvable domain and exposed RabbitMQ port for the server running this Docker environment)! The internal broker URLs shown above are only valid within the Docker environment.

Configuration¶

Repository Volumes¶

There are two volume mounts available: /config and /repos.

These are only required if using tools that require full repository access.

/config is a directory that may contain any of the following files:

repositories.json – A list of repositories Review Bot can pull from
servers.json – A list of Review Board servers that Review Bot can query for available repositories

Broker Configuration¶

The URL for the broker must be provided by passing BROKER_URL.

For RabbitMQ, this is in the form of: amqp://<username>:<password>@<host>/<queue-vhost>.

See Broker URL for more information.

Custom Review Bot Images¶

You can build your own custom Review Bot images featuring just the tools you need to install:

Determine the tools you want to use. The following tool IDs can be specified:
- checkstyle
- clang
- cppcheck
- cpplint
- doc8
- fbinfer
- flake8
- gofmt
- gotool
- jshint
- pmd
- pycodestyle
- pydocstyle
- pyflakes
- rubocop
- rustfmt
- shellcheck
Note

Built-in tools will always be installed. Currently, that just includes Secret Scanner.

Create a Dockerfile containing the following:

# Specify your tool IDs from above here:
ARG TOOLS="tool1 tool2 tool3"

FROM beanbag/reviewbot:<version>

Build the image:
```
$ docker build -t my-reviewbot .
```
See the docker build documentation for more information on this command.

Launch a container from your new image:

$ docker run \
    -P \
    --name ... \
    -v ... \
    -e ... \
    my-reviewbot