The sturdy d'ploy-mate

T'arrust me, bucko. I can do it.

This little gem can be added to your ruby-projects in order to set up a working ruby-server and deploy to it.

The resulting server will work with this setup:

NGINX Webserver (http://nginx.org/)
Unicorn Rack App Server (http://unicorn.bogomips.org/)
Bluepill Process Monitoring (https://github.com/bluepill-rb/bluepill)
RVM as user-install (http://rvm.io)
[Optional] ImageMagick (http://www.imagemagick.org)

Other stuff taken care of:

Logration
Automatic updates
NTP

You can choose a Database-Engine:

MySQL
Postgres

If you are not using Amazon AWS as a host: The GEM needs a working Amazon AWS-style ubuntu-user on the system in order to work properly. It will create one, if needed.

It uses the following Capistrano-roles to divide the installed components:

web: Machines with this role run NGINX as proxy
app: Machines with this role run the Ruby-webapp
search: [OPTIONAL] Machines with this role run ElasticSearch
cronjobs: [OPTIONAL] For environments where whenever should manage/run cronjobs

Changelog

2015-10-17: Added support for puma as application server
2015-10-12: Support additional linked-directories
2015-08-27: Small fix for nginx-config regarding gzipping of svgs
2015-08-14: Important small fix for nginx-config regarding ssl
2015-08-07: Small fix for rvm-install
2015-08-06: Support for optional sidekiq
2015-06-23: Support for optional elasticsearch
2015-06-23: Support for optional whenever
2015-06-23: Support for optional imagemagick
2015-06-23: Support creation of working ubuntu user. Error out if another user than ubuntu is used.
2015-06-22: Support for choosing your Ruby-version when creating the Capfile. Suggestions come from .ruby-version and Gemfile.
2015-04-29: Load custom rake tasks from lib/capistrano/tasks directory. You need to run the generator rake deploy_mate:install again or add Dir.glob('lib/capistrano/tasks/*.rake').each { |r| import r } to your Capfile.

Installation

Add this to your project's Gemfile:

gem 'deploy_mate'

and run bundle install.

After installing the gem you need to generate the files needed to be able to deploy with capistrano.

Add this at the end of your project's Rakefile:

load 'deploy_mate/tasks.rake'

then run

rake deploy_mate:install

Follow the instructions.

This will create the following files for you:

Capfile
config/deploy.rb
config/deploy/<your-stage>.rb

When done: Remove load 'deploy_mate/tasks.rake' from your Rakefile. It is not needed anymore and will otherwise only cause problems.

Updating the gem

Should you need to update your deploy_mate-version (e.g. because somebody fixed a bug in the gem), run:

bundle update deploy_mate

This will bump you up to the latest repo-version.

Setting up a server

Spawn yourself a basic Ubuntu 14 at the provider of your choice.
Create a working SSH-configuration for that server and try it our using ssh <your-server-name>
Run cap <your-stage> machine:init to install the needed packages.
Run cap <your-stage> machine:setup the setup all needed configuration-files on the server
Run cap <your-stage> deploy and be done.