Fetching contributors…
Cannot retrieve contributors at this time
179 lines (116 sloc) 6.15 KB

Asset Sync

Synchronises Assets between Rails and S3.

Asset Sync is built to run with the new Rails Asset Pipeline feature of Rails 3.1. After you run bundle exec rake assets:precompile your assets will be synchronised to your S3 bucket, optionally deleting unused files and only uploading the files it needs to.

This was initially built and is intended to work on Heroku


We are currently trying to talk with Heroku to iron these out.

  1. Will not work on heroku on an application with a RAILS_ENV configured as anything other than production
  2. Will not work on heroku using ENV variables with the configuration as described below, you must hardcode all variables


When you see rake assets:precompile during deployment. Herok is actually running something like

env RAILS_ENV=production DATABASE_URL=scheme://user:pass@ bundle exec rake assets:precompile 2>&1

This means the RAILS_ENV you have set via heroku:config is not used.

Workaround: you could have just one S3 bucket dedicated to assets and ensure to set keep the existing remote files

AssetSync.configure do |config|
  config.fog_directory = 'app-assets'
  config.existing_remote_files = "keep"

2. ENV varables not available

Currently when heroku runs rake assets:precompile during deployment. It does not load your Rails application's environment config. This means using any ENV variables you could normally depend on are not available.

Workaround: you could just hardcode your AWS credentials in the initializer or yml

AssetSync.configure do |config|
  config.aws_access_key_id = 'xxx'
  config.aws_secret_access_key = 'xxx'
  config.fog_directory = 'mybucket'


Add the gem to your Gemfile

gem "asset_sync"

Generate the rake task and config files

rails g asset_sync:install

If you would like to use a YAML file for configuration instead of the default (Rails Initializer) then

rails g asset_sync:install --use-yml


Configure config/environments/production.rb to use Amazon S3 as the asset host and ensure precompiling is enabled.

# config/environments/production.rb
config.action_controller.asset_host = do |source, request|
  request.ssl? ? "https://#{ENV['FOG_DIRECTORY']}" : "http://#{ENV['FOG_DIRECTORY']}"

We support two methods of configuration.

  • Rails Initializer
  • A YAML config file

Using an Initializer is the default method and is best used with environment variables. It's the recommended approach for deployments on Heroku.

Using a YAML config file is a traditional strategy for Capistrano deployments. If you are using Moonshine (which we would recommend) then it is best used with shared configuration files.

The recommend way to configure asset_sync is by using environment variables however it's up to you, it will work fine if you hard code them too. The main reason is that then your access keys are not checked into version control.

Initializer (config/initializers/asset_sync.rb)

The generator will create a Rails initializer at config/initializers/asset_sync.rb.

AssetSync.configure do |config|
  config.aws_access_key_id = ENV['AWS_ACCESS_KEY_ID']
  config.aws_secret_access_key = ENV['AWS_SECRET_ACCESS_KEY']
  config.fog_directory = ENV['FOG_DIRECTORY']
  # config.fog_region = 'eu-west-1'
  config.existing_remote_files = "keep"

YAML (config/asset_sync.yml)

If you used the --use-yml flag, the generator will create a YAML file at config/initializers/asset_sync.rb.

defaults: &defaults
  aws_access_key_id: "<%= ENV['AWS_ACCESS_KEY_ID'] %>"
  aws_secret_access_key: "<%= ENV['AWS_SECRET_ACCESS_KEY'] %>"
  # You may need to specify what region your S3 bucket is in
  # fog_region: "eu-west-1"

  <<: *defaults
  fog_directory: "rails-app-development"
  existing_remote_files: keep # Existing pre-compiled assets on S3 will be kept

  <<: *defaults
  fog_directory: "rails-app-test"
  existing_remote_files: keep

  <<: *defaults
  fog_directory: "rails-app-production"
  existing_remote_files: delete # Existing pre-compiled assets on S3 will be deleted

Environment Variables

Add your Amazon S3 configuration details to heroku

heroku config:add AWS_ACCESS_KEY_ID=xxxx
heroku config:add AWS_SECRET_ACCESS_KEY=xxxx
heroku config:add FOG_DIRECTORY=xxxx

Or add to a traditional unix system

export AWS_ACCESS_KEY_ID=xxxx
export FOG_DIRECTORY=xxxx

Available Configuration Options

  • aws_access_key_id: your Amazon S3 access key
  • aws_secret_access_key: your Amazon S3 access secret
  • aws_region: the region your S3 bucket is in e.g. eu-west-1
  • existing_remote_files: what to do with previously precompiled files, options are keep or delete

Amazon S3 Multiple Region Support

If you are using anything other than the US buckets with S3 then you'll want to set the region. For example with an EU bucket you could set the following with YAML.

  # ...
  aws\_region: 'eu-west-1'

Or via the initializer

AssetSync.configure do |config|
  # ...
  config.fog_region = 'eu-west-1'

Rake Task

A rake task is installed with the generator to enhance the rails precompile task by automatically running after it:

# lib/tasks/asset_sync.rake
Rake::Task["assets:precompile"].enhance do


  1. Add some before and after filters for deleting and uploading
  2. Support more cloud storage fog_providers
  3. Better test coverage


Have borrowed ideas from:


MIT License. Copyright 2011 Rumble Labs Ltd.