Table of Content:

• Ansible Role: RabbitMQ Cluster
• Requirements
• Must-have Role Variables
• Optional Role Variables and Defaults
• Example
• License
• Author Information

Ansible Role: RabbitMQ Cluster

An Ansible Role, that installs a RabbitMQ multi-node cluster.

This project is inspired by alexey-medvedchikov/ansible-rabbitmq.

It supports Debian/Ubuntu only.

Note from 2020.12.31: this project is from 3 years ago and I did not maintain/update on a regular basis. For more recent version which works out of the box, please search for other projects. Thanks.

Major changes and differences:

• TLS/SSL
• high availability queues
• Simple template and config
• More comments and more readable code

Requirements

None, but note that hostname -f should work on every node, and this is especially worth paying some attention if you work with AWS EC2 instances, since in aws the host name is like ip-xxx-xxx-xxx-xxx.eu-central-1.compute.internal.

Must-have Role Variables

Must have variables are listed below:

rabbitmq_cluster_master: hostname

This parameter is used to determine which node is the master, because some tasks should only be run on master or slave, like slave joins into a cluster.

This role uses ansible fqdn equals rabbitmq_cluster_master or not to determine if it's master or slave.

Optional Role Variables and Defaults

All the other variables are optional and listed below along with default values (see defaults/main.yml):

update_hosts: false

Whether you need to update hosts file or not, default false. This is useful when you are using AWS EC2 instance, whose default hostname is too long and doesn't have a meaning, like "ip-10-101-50-12.eu-central-1.compute.internal", but you want to use something shorter and meaningful as hostname. In this case you need to set this variable to true in order to update the hosts file, and you need to define a variable named "rabbitmq_hosts", with the following format:

rabbitmq_hosts: |
  node-1-ip node-1-FQDN
  node-2-ip node-2-FQDN

example:

rabbitmq_hosts: |
  10.0.0.10 eu-central-1-mq-master   (whatever the command `hostname -f` outputs on this host)
  10.0.0.11 eu-central-1-mq-slave-01 (whatever the command `hostname -f` outputs on this host)

.

rabbitmq_create_cluster: yes

To use cluster or not. Default yes, which means slave will join the master as a cluster.

rabbitmq_erlang_cookie: WKRBTTEQRYPTQOPUKSVF

RabbitMQ nodes and CLI tools (e.g. rabbitmqctl) use a cookie to determine whether they are allowed to communicate with each other. For two nodes to be able to communicate they must have the same shared secret called the Erlang cookie.

For more info see: https://www.rabbitmq.com/clustering.html

rabbitmq_use_longname: 'false'

When set to true this will cause RabbitMQ to use fully qualified names to identify nodes. This may prove useful on EC2. Note that it is not possible to switch between using short and long names without resetting the node.

For more info see: https://www.rabbitmq.com/configure.html#define-environment-variables

rabbitmq_logrotate_period: weekly
rabbitmq_logrotate_amount: 20

Log rotation settings.

rabbitmq_ulimit_open_files: 4096

The main setting that needs adjustment is the max number of open files, also known as ulimit -n. The default value on many operating systems is too low for a messaging broker (eg. 1024 on several Linux distributions). RabbitMQ recommends allowing for at least 65536 file descriptors for user rabbitmq in production environments. 4096 should be sufficient for most development workloads.

More info at: https://www.rabbitmq.com/install-debian.html

rabbitmq_tls_port: 5671
rabbitmq_amqp_port: 5672
rabbitmq_epmd_port: 4369
rabbitmq_node_port: 25672

RabbitMQ default ports.

rabbitmq_plugins:
  - rabbitmq_management
  - rabbitmq_management_agent
  - rabbitmq_shovel
  - rabbitmq_shovel_management

Plugins installed by default, mainly for HTTP API monitor.

enable_tls: false

Enable TLS/SSL or not.

tls_only: false

If true, only TLS port is open; default amqp port 5672 will be disabled.

tls_verify: "verify_none"
tls_fail_if_no_peer_cert: false

Settings for TLS. More info at: https://www.rabbitmq.com/ssl.html

cacertfile: ""
cacertfile_dest: "/etc/rabbitmq/cacert.pem"

certfile: ""
certfile_dest: "/etc/rabbitmq/cert.pem"

keyfile: ""
keyfile_dest: "/etc/rabbitmq/key.pem"

If TLS is enabled, you need to specify cacert/certfile/keyfile location, and where to put them on the server.

backup_queues_in_two_nodes: false

By default, queues within a RabbitMQ cluster are located on a single node (the node on which they were first declared). Queues can optionally be made mirrored across all nodes, or exactly N number of nodes. By enabling this variable to true, there will be 1 queue master and 1 queue mirror. If the node running the queue master becomes unavailable, the queue mirror will be automatically promoted to master.

More info see: https://www.rabbitmq.com/ha.html

Example

Playbook:

---
- hosts: mq-cluster
  become: yes
  become_user: root
  roles:
    - rabbitmq

Group vars file 'mq-cluster':

rabbitmq_cluster_master: mq-cluster-master
update_hosts: true
rabbitmq_hosts: |
  10.0.0.10 mq-cluster-master
  10.0.0.11 mq-cluster-slave-01
enable_tls: true
cert_dir: "{{ playbook_dir }}/../common_files/ssl"
cacertfile: "{{ cert_dir }}/cacert.pem"
certfile: "{{ cert_dir }}/cert.pem"
keyfile: "{{ cert_dir }}/key.pem"
backup_queues_in_two_nodes: true

Inventory file:

[mq-cluster]
10.0.0.10
10.0.0.11

License

MIT / BSD

Author Information

This role was created between late 2017 and early 2018 by [Tiexin Guo].