RubyRescue

Ruby Rescue uses unicorn for all of our deployed Ruby applications. Chris Wanstrath promoted it in a blog post on github, and we decided soon after to give it a shot. Because it’s based on mongrel, it takes a reliable and well-tested app server and bolts a ruby layer on top that handles restarts, adding and removing workers, and uses unix sockets to wait for requests.

There’s a good article by Ryan Tomayko on Unicorn and the reasons why it’s “Unixyness” is a feature. If you’re a long-time unix user, you won’t find much to be revolutionary. If you’re a web-developer without a CS-background, he covers how unix fundamentals make Unicorn more reliable.

Before Unicorn we were running Passenger and apache, and we decided to switch to unicorn and nginx at the same time. It takes a bit more work to setup unicorn than passenger, but we’ve found the reliability to be worth it.

• gem install unicorn
• create a unicorn.rb file like this

rails_env = ENV['RAILS_ENV'] || 'production'

worker_processes 3

preload_app true

timeout 75

socket_path = '/var/www/APPLICATION/shared/sockets/unicorn.sock'
pid_path = '/var/www/APPLICATION/shared/pids/unicorn.pid'

listen socket_path, :backlog => 2048
pid pid_path

# http://www.rubyenterpriseedition.com/faq.html#adapt_apps_for_cow
if GC.respond_to?(:copy_on_write_friendly=)
  GC.copy_on_write_friendly = true
end

before_fork do |server, worker|
  old_pid = RAILS_ROOT + '/tmp/pids/unicorn.pid.oldbin'
  if File.exists?(old_pid) && server.pid != old_pid
    begin
      Process.kill("QUIT", File.read(old_pid).to_i)
    rescue Errno::ENOENT, Errno::ESRCH
    end
  end
end

after_fork do |server, worker|
  ActiveRecord::Base.establish_connection

  begin
    uid, gid = Process.euid, Process.egid
    user, group = 'admin', 'admin'
    target_uid = Etc.getpwnam(user).uid
    target_gid = Etc.getgrnam(group).gid
    worker.tmp.chown(target_uid, target_gid)
    if uid != target_uid || gid != target_gid
      Process.initgroups(user, target_gid)
      Process::GID.change_privilege(target_gid)
      Process::UID.change_privilege(target_uid)
    end
  rescue => e
    raise e
  end
end

• Your deploy.rb will get a bit more complicated as it needs to send the proper signal to the Unicorn master to signal it to reload the workers.

namespace :deploy do
  task :stop, :roles => :app do
  run "cd #{current_path} && kill -QUIT `cat tmp/pids/unicorn.pid`"
end 

task :start, :roles => :app do
  run "cd #{current_path} && /opt/ruby/bin/unicorn_rails -c config/unicorn.rb -E production -D"
end

desc "restart unicorn"
task :restart, :roles => :web do
  run "cd #{current_path}; [ -f tmp/pids/unicorn.pid ] && kill -USR2 `cat tmp/pids/unicorn.pid` || /opt/ruby/bin/unicorn_rails -c config/unicorn.rb -E production -D"
end