Capistrano recipes for Rails

We deploy a lot of Rails applications and our developers have to solve similar problems each time during the deployment: how to run workers, how to generate crontab, how to precompile assets faster and so on. This collection of recipes helps us to solve them.

Recipe use:

foreman + foreman_export_runitu to generate runit scripts with the Procfile
whenever to generate crontab
nginx as proxy server

It also consider that you use rbenv on the server.

##Installation

gem 'capistrano_rails_recipes', :require => false

Capfile example:

require "capistrano_rails_recipes/capistrano"

set :repository, "git@github.com:..."
set :application, "my_application"
set :user, 'deploy'
set :branch, 'master'
set :deploy_to, "/path/to/app"
set :use_sudo, false

server 'web.example.com', :web, :app, :worker, :crontab
role :db, 'web.example.com', primary: true

OR last two line can be expressed like

task :production do
  role  :web,     "web.example.com"
  role  :app,     "app.example.com"
  role  :crontab, "app.example.com"
  role, :db,      "db.example.com", :primary => true
  role, :worker,  "workers.example.com"
end

task :staging do
  server "stage.example.com", :web, :app, :crontab, :db, :worker
end

As you can see, we use use roles to bind the tasks, and there are some additions to roles and additional roles:

deploy:setup creates all necessary folders and symlinks to files:

config/nginx.conf                     -> /opt/nginx/conf/sites-enabled/my_application
creates deploy_to/shared/config folder
creates deploy_to/services folder
copies config/database.example.yml    -> shared/config/database.yml

Run deploy:setup before your first deploy. See below how to configure services.

web compiles assets if content of app/assets was changed since last deploy (add FORCE=1 to force the assets compilation)

app all files from shared/config is being symlinked to current/config like:

shared/config/database.yml            -> current/config/database.yml
shared/config/settings/production.yml -> current/config/settings/production.yml

crontab generates crontab with whenever gem, only if the content of config/schedule.rb was changed (add FORCE=1 to force the crontab generation)

db run migrations only if db/migrate was changed (add FORCE=1 to force migrations or SKIP_MIGRATION=1 to skip them)

worker Procfile exports runit configs to deploy_to/application/services

On deploy:restart runit workers is being restarted.

You can use some extra cap tasks:

rails:console to launch rails console on remote server
rails:dbconsole to launch rails dbconsole on remote server
login to open SSH session under user deploy and switch catalog to Capistrano's current_path

Important

To run succesfully together with system wide rbenv, all you tasks in Procfile must be started with rbenv exec

##Configuring services

To run services just create Procfile in root of your app.

Procfile example:

sidekiq: rbenv exec bundle exec sidekiq -L sidekiq.log
web: rbenv exec bundle exec unicorn -c config/unicorn.rb -E production

Also you need to configure runit to monitor deploy_to/services folder.

##Nginx config file example for Unicorn

upstream unicorn_my_application {
  server unix:/tmp/unicorn.my_application.sock fail_timeout=0;
}

server {
  listen 80;
  server_name web.example.com;
  root deploy_to/current/public;

  location ^~ /assets/ {
    gzip_static on;
    expires max;
    add_header Cache-Control public;
  }

  try_files $uri/index.html $uri @unicorn;
  location @unicorn {
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header Host $http_host;
    proxy_redirect off;
    proxy_pass http://unicorn_my_application;
  }

  error_page 500 502 503 504 /500.html;
  client_max_body_size 4G;
  keepalive_timeout 10;
}

Change deploy_to and my_application to your actual values.

##Unicorn config file example

root = "deploy_to/current"
working_directory root
pid "#{root}/tmp/pids/unicorn.pid"
stderr_path "#{root}/log/unicorn.log"
stdout_path "#{root}/log/unicorn.log"

listen "/tmp/unicorn.my_application.sock"
worker_processes 5
timeout 30

Change deploy_to and my_application to your actual values.

##Capistrano

Default variables:

logger.level                   = Capistrano::Logger::DEBUG
default_run_options[:pty]      = true
ssh_options[:forward_agent]    = true
set :bundle_cmd,                 "rbenv exec bundle"
set :bundle_flags,               "--deployment --quiet --binstubs --shebang ruby-local-exec"
set :rake,                       -> { "#{bundle_cmd} exec rake" }
set :keep_releases,              7
set :scm,                        "git"
set :user,                       "deploy"
set :deploy_via,                 :unshared_remote_cache
set :copy_exclude,               [".git"]
set :repository_cache,           -> { "#{deploy_to}/shared/#{application}.git" }
set :normalize_asset_timestamps, false

##Bonus track

To enable silent mode, add ENV['CAP_SILENT_MODE'] before the require 'capistrano_evrone_recipes/capistrano' in your Capfile

capistrano_rails_recipes

Runtime

Capistrano recipes for Rails