Capistrano::Bundler
Bundler specific tasks for Capistrano v3:
$ cap production bundler:install
It also prefixes certain binaries to use bundle exec
.
Installation
Add these lines to your application's Gemfile [Recommended]:
gem 'capistrano', '~> 3.6'
gem 'capistrano-bundler', '~> 2.0'
And then execute:
$ bundle
Or install it yourself as:
$ gem install capistrano-bundler
Usage
Require in Capfile
to use the default task:
require 'capistrano/bundler'
The task will run before deploy:updated
as part of Capistrano's default deploy, or can be run in isolation with cap production bundler:install
.
In order for Bundler to work efficiently on the server, its project configuration directory (<release_path>/.bundle/
) should be persistent across releases.
You need to add it to the linked_dirs
Capistrano variable:
Capistrano 3.5+:
append :linked_dirs, '.bundle'
Capistrano < 3.5:
set :linked_dirs, fetch(:linked_dirs, []) << '.bundle'
It will still work fine with non-persistent configuration directory, but then it will have to re-resolve all gems on each deploy.
By default, the plugin adds bundle exec
prefix to common executables listed in bundle_bins
option. This currently applies for gem
, rake
and rails
.
You can add any custom executable to this list:
set :bundle_bins, fetch(:bundle_bins, []).push('my_new_binary')
Configurable options:
set :bundle_roles, :all
set :bundle_config, { deployment: true }
set :bundle_servers, -> { release_roles(fetch(:bundle_roles)) }
set :bundle_binstubs, -> { shared_path.join('bin') }
set :bundle_binstubs_command, :install
set :bundle_gemfile, -> { release_path.join('MyGemfile') }
set :bundle_path, -> { shared_path.join('bundle') }
set :bundle_without, %w{development test}.join(':')
set :bundle_flags, '--quiet'
set :bundle_env_variables, {}
set :bundle_clean_options, ""
set :bundle_check_before_install, true
You can parallelize the installation of gems with bundler's jobs feature.
Choose a number less or equal than the number of cores your server.
set :bundle_jobs, 8
To generate binstubs on each deploy, set :bundle_binstubs
path:
set :bundle_binstubs, -> { shared_path.join('bin') }
In the result this would execute the following bundle commands on all servers
(actual paths depend on the real deploy directory):
$ bundle config --local deployment true
$ bundle config --local gemfile /my_app/releases/20130623094732/MyGemfile
$ bundle config --local path /my_app/shared/bundle
$ bundle config --local without "development test"
$ bundle install --quiet --binstubs /my_app/shared/bin
If any option is set to nil
it will be excluded from the final bundle commands.
If you want to clean up gems after a successful deploy, add after 'deploy:published', 'bundler:clean'
to config/deploy.rb.
Downsides to cleaning:
- If a rollback requires rebuilding a Gem with a large compiled binary component, such as Nokogiri, the rollback will take a while.
- In rare cases, if a gem that was used in the previously deployed version was yanked, rollback would entirely fail.
If you're using Bundler >= 2.1 and you are generating binstubs, you can configure capistrano-bundler to use the newer
bundle binstubs
command. This will avoid the deprecation warning that you'd otherwise get when using bundle install
to generate binstubs:
set :bundle_binstubs_command, :binstubs
Environment Variables
The bundle_env_variables
option can be used to specify any environment variables you want present when running the bundle
command:
set :bundle_env_variables, { nokogiri_use_system_libraries: 1 }
Contributing
- Fork it
- Create your feature branch (
git checkout -b my-new-feature
) - Commit your changes (
git commit -am 'Add some feature'
) - Push to the branch (
git push origin my-new-feature
) - Create new Pull Request