Migrake

A simple Rake extension that lets you define tasks that are divided into multiple smaller "versions". When you run the task in a given host, it will only run the "versions" that haven't yet been run in that host. Think something similar to ActiveRecord::Migration for Rake tasks.

This is specially useful to define tasks that should be run on deploy, in order to do some sort of house cleaning that should be run once. For example, some complex data migrations, or re-processing your user avatars because now the UI shows them in a different size.

The important thing is these tasks should only be run once when you deploy to a given environment, and they should be run immediately.

An example

In lib/tasks/migrake.rake:

require "migrake"

migrake Set.new([
  "data:set_default_user_state",
  "twitter:purge_cache",
  # etc, etc
])

This would define a task called :migrake that, when called, will invoke the tasks in that set unless they have been invoked before. This means that after each deploy you can run this task (with a capistrano hook, for example) and any tasks that need to be run in this environment will be run.

How it works

This will keep a file named MIGRAKE_STATUS, that will contain the name of the tasks that were ran in the past. Whenever you run rake migrake migrake will check that file, see which tasks in the set aren't in it, and will run those tasks (no order is guaranteed, if you need tasks to run in order, define the dependencies in the task themselves.)

Overriding the file where tasks are stored

The file will be located, by default, at whichever directory rake is invoked from. In order to change the path to the file, you can do this:

Migrake.store = Migrake::FileSystemStore.new("./config/migrake")

This must be called in your Rakefile before you define the tasks to be run by migrake. Like this:

require "migrake"

Migrake.store = Migrake::FileSystemStore.new("./config/migrake")

migrake Set.new([
  # ...
])

As an alternative, you can make MIGRAKE_STATUS_DIR available to your environment, and migrake will use a MIGRAKE_STATUS file in that directory.

MIGRAKE_STATUS_DIR=./some/path bundle exec rake migrake

This is particularly useful to use with capistrano, since you probably want to use cap's shared directory to store your migrake file. For example:

namespace :deploy do
  task :migrake do
    run "MIGRAKE_STATUS_DIR=#{shared_path} bundle exec rake migrake"
  end
end

after "deploy:migrations", "deploy:migrake"

Heroku and other environments where you can't easily write to the filesystem

The FileSystemStore is easily swappable. For the particular case of Heroku and similar servers we provide the Migrake::SQLStore gem, that implements the storage API using a database table.

Writing a new storage engine should be trivial. Use either FileSystemStore or SQLStore as an example.

Bootstrapping a new environment

When you bootstrap a new environment you don't need to run migrake tasks that have been already run. For this, when the migrake method is invoked we also define a migrake:ready task, that forces all tasks defined in the migrake set into the MIGRAKE_STATUS file.

This way you can just run that when you bootstrap the environment and then keep running rake migrake when you deploy.

License

(The MIT License)

Copyright (c) 2012 Nicolas Sanguinetti, with the support of Cubox

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.