Deploying a Ruby application on a server of your own can get challenging when you wish to make the deployment procedure simple.

Recently, I got myself a Digital Ocean account. My primary motivation of shifting to a VPS was to learn more about deploying applications and to be able to ship out my side projects.

It would hardly take you a couple of minutes to get a droplet up and running.

I quickly created a GitHub repository called hello and wrote a tiny Sinatra application. After this, I created a user account on my server for handling the deployments.

My first choice for a deployment tool was Capistrano and I setup my application to use it. After fiddling with Capistrano for a few days, I realised that SSH Kit doesn’t really work that well for me. Having been used to a Grunt based deployment setup at my present organisation, Capistrano’s in-built tasks did not suit my needs. At present, I wish to have more control over the deployment procedure. Since, I had heard so much praise about the tool and because of the abundant documentation available online, I decided to give it some more time. Even after a few more tries, it didn’t really work out.

Initially, when I was looking around for deployment tools, I had come across Mina, but for some strange reason I dismissed it. After my repeated failures with Capistrano, I decided to use Mina. What Mina does is that it creates a single batch script which it then sends to the server and executes. According to me, this is a fantastic technique because it reduces the number of requests made during the procedure. Also, it allows for the usage of other commands straight away.

My current deploy script is pretty neat:

require 'mina/bundler'
require 'mina/git'
require 'mina/rbenv'
require 'configatron'

require './config/settings.rb'

set :domain,
set :deploy_to, configatron.server.deploy_to

set :repository, configatron.git.repository
set :branch, configatron.git.branch

set :user, configatron.ssh.user
set :forward_agent, configatron.ssh.forward_agent
set :identity_file, configatron.ssh.identity_file

set :shared_paths, ['log', 'tmp']

task :environment do
  invoke :'rbenv:load'

task :setup => :environment do
  queue! %[mkdir -p "#{deploy_to}/shared/log"]
  queue! %[chmod g+rx,u+rwx "#{deploy_to}/shared/log"]

  queue! %[mkdir -p "#{deploy_to}/shared/tmp"]
  queue! %[chmod g+rx,u+rwx "#{deploy_to}/shared/tmp"]

desc "Deploys the current version to the server."
task :deploy => :environment do
  deploy do
    invoke :'git:clone'
    invoke :'deploy:link_shared_paths'
    invoke :'bundle:install'

    to :launch do
      invoke :restart

desc "Restart the server."
task :restart => :environment do
  in_directory "#{deploy_to}/#{current_path}" do
    queue "bundle exec thin -R -p 4567 -d restart"

desc "Start the server."
task :start => :environment do
  in_directory "#{deploy_to}/#{current_path}" do
    queue "bundle exec thin -R -p 4567 -d start"

desc "Stop the server."
task :stop => :environment do
  in_directory "#{deploy_to}/#{current_path}" do
    queue "bundle exec thin -R -p 4567 -d stop"

desc "Report server process id"
task :info => :environment do
  in_directory "#{deploy_to}/#{shared_path}" do
    queue "print 'Server running with pid:' `cat tmp/pids/`"