Ansible Unicorn/RVM setup

This Ansible role installs a service for Unicorn, with RVM and multiple application support

Requirements

Expects rvm to be installed system-wide

Role Variables

• rvm_root defaults to /usr/local/rvm

• rvm_exe defaults to /usr/local/rvm/bin/rvm

• rails_apps defaults to []

  Each rails_apps entry is a dict with the following options

  • name (eg my_app, required)

  • ruby_version (eg ruby-2.1.1, required)

  • env defaults to production

  • root defaults to /var/www/{{ name }}/current (Capistrano compatible)

  • config is used for db and secrets, defaults to {{ root }}/config if root is specified, otherwise to /var/www/{{ name }}/shared/config

  • db is a dict for generating a database.yml file, with options:

    • adapter defaults to postgresql
    • host defaults to localhost
    • database defaults to app name
    • username defaults to app name
    • password defaults to app name
    • pool defaults to 5
    • timeout defaults to 5000

  • secrets is a dict and all its keys are converted to secrets.yml

Example Playbook

The role could be included in a playbook as follows:

---
-hosts: application
  roles:
    - role: unicorn-rvm
      rails_apps:
        - { name: 'my_app1', ruby_version: 'ruby-1.9.3' }
        - { name: 'my_app2', ruby_version: 'ruby-2.1.1', root: '/var/test_apps/app2', env: staging }
        - name: 'my_app3'
          ruby_version: 'ruby-2.1.1'
          db:
            password: topsecret
          secrets:
            secret_key_base: SuperSecretHexString

If the init script is called without any config parameters, it will attempt to run the init command for all unicorn configurations listed in /etc/unicorn/*.conf

/etc/init.d/unicorn start # starts all unicorns

If you specify a particular config, it will only operate on that one

/etc/init.d/unicorn start my_app

Notes

This role does not deploy the actual application, it assumes that this will be done in another role or using a deployment tool such as Capistrano.

However you deploy, keep in mind that this setup expects each application to use an rvm gemset with the name of the application.

So for example, for a Capistrano 3 deployment of an app called 'my_app' under ruby 2.1.1, you should require 'capistrano/rvm' and set the following in your Capistrano configuration:

set :rvm_ruby_version, '2.1.1@my_app'

If you use this role to generate database.yml and secrets.yml, it is advised to store the keys in an ansible vault file

License

MIT