Clustering RabbitMQ Docker containers

This repository accompanies a blog article/YouTube video I created to walk people through the setup of a RabbitMQ cluster on their machine. Go to my website to learn more about the setup:

Set up a RabbitMQ cluster on your laptop using Docker

Using docker run

In this section you will learn the commands used to launch a RabbitMQ cluster using the regular docker engine run command.

1. Start master node:

docker run -d \
    --name="rabbit1" \
    --hostname="rabbit1"\
    -e RABBITMQ_ERLANG_COOKIE="secret string" \
    -e RABBITMQ_NODENAME="rabbit1" \
    --volume=(pwd)/rabbitmq.config:/etc/rabbitmq/rabbitmq.config \
    --volume=(pwd)/definitions.json:/etc/rabbitmq/definitions.json \
    --publish="4369:4369" \
    --publish="5671:5671" \
    --publish="5672:5672" \
    --publish="15671:15671" \
    --publish="15672:15672" \
    --publish="25672:25672" \
    rabbitmq:3-management

2. Start slave #1:

docker run -d \
    --name="rabbit2" \
    --hostname="rabbit2"\
    -e RABBITMQ_ERLANG_COOKIE="secret string" \
    -e RABBITMQ_NODENAME="rabbit2" \
    --volume=(pwd)/rabbitmq.config:/etc/rabbitmq/rabbitmq.config \
    --volume=(pwd)/definitions.json:/etc/rabbitmq/definitions.json \
    --link="rabbit1:rabbit1" \
    rabbitmq:3-management

3. Start slave #2:

docker run -d \
    --name="rabbit3" \
    --hostname="rabbit3"\
    -e RABBITMQ_ERLANG_COOKIE="secret string" \
    -e RABBITMQ_NODENAME="rabbit3" \
    --volume=(pwd)/rabbitmq.config:/etc/rabbitmq/rabbitmq.config \
    --volume=(pwd)/definitions.json:/etc/rabbitmq/definitions.json \
    --link="rabbit1:rabbit1" \
    --link="rabbit2:rabbit2" \
    rabbitmq:3-management

4. View container logs individually

docker logs -f <rabbit#>

This will display the logs for the chosen container, and follow them just like tail -f /log/path would do.

5. Run producer / consumer

The producer / consumer scripts were created as simple Node.js scripts so they can be executed using regular bash script execution syntax.

Using docker-compose

This section details how to start the whole cluster using docker-compose and a YAML definition file

1. Create a network shared by all containers

docker network create rabbitmq-cluster

2. Start cluster:

docker-compose up -d

3. View logs for all containers

docker-compose logs -f

3. Run producer / consumer for testing

Same as in the section detailing docker run you can launch the consumer/producer as regular shell scripts.

Create a distributed local cluster

To be continued

Deploying to AWS

To be continued

Deploying to Microsoft Azure

To be continued

Tools

Very useful tool for simulating message circulation in your RabbitMQ setup.

http://tryrabbitmq.com/