Fork me on GitHub

Capistrano is an open source tool for running scripts on multiple servers. It’s primary use is for easily deploying applications. While it was built specifically for deploying Rails apps, it’s pretty simple to customize it to deploy other types of applications.
capifony is a deployment recipes collection that works with both symfony and Symfony applications.

1. Install script

First, install capifony itself with RubyGems:

gem install capifony

2. Setup project

Setup initial deployment configuration for your project:

cd path/to/your/project
capifony .

This will create a Capfile in your project root, and a deploy.rb config file in config/ (for symfony 1.x projects) or in app/config/ (for Symfony2 projects).

3. Configure

Fill up your config/deploy.rb (or app/config/deploy.rb) with your server connection data. But first, you must choose your deployment strategy.

a) deployment → scm → production

The first strategy involves deployment to a production server via an intermediate git repository server. In this case, the production server must have access to your Git repository (remote or not) and be able to issue a “pull” from it. You must have ssh access to the production server from wherever you’re deploying from:

# deploy.rb

set   :application,   "My App"
set   :deploy_to,     "/var/www/"
set   :domain,        ""

set   :scm,           :git
set   :repository,    ""

role  :web,           domain
role  :app,           domain, :primary => true

set   :use_sudo,      false
set   :keep_releases, 3

In this case, on every cap deploy, capifony will:

NOTE : By default capifony doesn’t clean releases. To execute the clean releases task, it’s necessary to set the keep_releases param and to add the task after end deploy in the deploy.rb file, like this: after "deploy", "deploy:cleanup"

If you don’t want to clone the whole repository on every deploy, you can set the :deploy_via parameter:

set :deploy_via, :remote_cache

In this case, a git repository will be kept on the server and Capifony will only fetch the changes since the last deploy.

b) deployment → production (via copy)

The second strategy involves deployment to a production server right from your deployment machine via a copy. In this case, the deployment server (which may just be your local computer) must have access to the git repository (remote or not), and be able to pull from it.

As vendors are still installed on the production server it is worth noting that if any of your vendors are remote private repositories you will need pull access for any of these repositories on your production server otherwise install via composer or bin/vendors will fail. You will need to use the following cookbook entry if you want to install vendors locally before deploy - Install vendors locally before deploy.

The deployment server must also have ssh access to the production server:

# deploy.rb

set   :application,   "My App"
set   :deploy_to,     "/var/www/"
set   :domain,        ""

set   :scm,           :git
set   :repository,    "file:///Users/deployer/sites/my-app"
set   :deploy_via,    :copy

role  :web,           domain
role  :app,           domain, :primary => true

set   :use_sudo,      false
set   :keep_releases, 3

In this case, on every cap deploy, capifony will:

Of course, copying the whole project on every deploy is very expensive and slow. Fortunately, you can optimize things with the capistrano_rsync_with_remote_cache gem:

gem install capistrano_rsync_with_remote_cache

Now, change your deployment strategy in deploy.rb:

set :deploy_via, :rsync_with_remote_cache

In this case, rsync will create a cache on your production server and will push only files that have changed between deploys.

4. Setup server

Now, you can start the deployment process! To get your server setup with the directory structure that Capistrano expects, cd to your local project directory and run:

cap deploy:setup

(You’ll only have to run this once!)

This command will create the following approximate directory structure on your server. The exact structure will depend on if you’re deploying a symfony 1.x or Symfony2 application:

`-- /var/www/
|-- current → /var/www/
|-- releases
|   `-- 20100512131539
|   `-- 20100509150741
|   `-- 20100509145325
`-- shared
   |-- web
   |    `-- uploads
   |-- log
   `-- config
       `-- databases.yml

The folders in the releases directory will be the actual deployed code, as timestamped directories. In a symfony 1.x application, for example, Capistrano symlinks your log & web/uploads directories from your app to the directories in the shared folder so that it doesn’t get erased when you deploy a new version of your code.

To quickly setup a new server, you can do:

cap deploy:setup

5. Deploy!

To deploy your application, simply run:

cap deploy

Depending on your setup, you may need to ssh onto your server to setup additional, shared files after your first deployment (e.g. app/config/parameters.yml if you’re using Symfony2 with the deployment recipe listed below).

Something went wrong???

cap deploy:rollback

Symfony2 Deployment

If you’re deploying a Symfony2 application, then this section is probably for you. This section explains how to configure capifony to deploy an application that uses the bin/vendors file to manage vendor libraries and the app/config/parameters.yml file for server-specific configuration (like database connection information).

First, add the following to your app/config/deploy.rb file so that the parameters.yml file is shared between all deployments:

set :shared_files,      ["app/config/parameters.yml"]

Next, share the vendor directory between all deployments to make deploying faster:

set :shared_children,     [app_path + "/logs", web_path + "/uploads", "vendor"]

To install your vendors, capifony will rely on bin/vendors by default. But the recommended dependency manager is now Composer. In order to use it, just add the following configuration:

set :use_composer, true

If you want to update your vendors, add the following parameter:

set :update_vendors, true

It will run composer.phar update if you use Composer, bin/vendors otherwise. Note that the bin/vendors can be configured using the :vendors_mode parameter to decide which action to run (upgrade, install, or reinstall).

The final step is to configure your app/config/parameters.yml file. The best way to do this is to create this file in shared folder on server manually:

ssh your_deploy_server
mkdir -p /var/www/
vim /var/www/

Once your parameters.yml file is correctly configured, you should be able to test your deployed application. On every subsequent deploy, that same app/config/parameters.yml file will be symlinked into your application, meaning you only need to configure after the initial deploy.

Useful Tasks

If you need to deploy and run your migrations you can call:

cap deploy:migrations

To run your test suite on the production server, just invoke:

cap deploy:test_all

You can invoke tasks/commands by calling:

cap symfony

If you want to see all available tasks, you can run:

cap -vT

Configuration References

Capistrano Configuration

Symfony Configuration

## Cookbook

To learn more about capifony, you can read these recipes:

Enabling/Disabling applications

Using the Multistage extension

Shared folders with symfony 1.x

Specifying Deployment Dependencies

Managing External Vendors with Composer

Setup an automatic upload of parameters file

Setup proper permissions

Speeding up deploy

Working with databases

Deployment in clustered machines

How to keep Symfony2 sessions after deploy

Install vendors locally before deploy

Known Issues

If you get the following error message sudo : no tty present and no askpass program specified, add this parameter:

default_run_options[:pty] = true

Found a typo? Something is wrong in this documentation? Just fork and edit it !