At Ombu Labs, we are big fans of Ruby on Rails and design patterns, especially convention over configuration! The beauty of Rails is that you can inherit a legacy project and easily find the different layers of code in different directories.
When it comes to database migrations the policy of Rails is very clear. It's all about altering the database structure with gradual migration files: "Migrations are a convenient way to alter your database schema over time in a consistent and easy way." (source)
But, what about data migrations? What's the best way to write, maintain, and run migrations that alter the data in your production database?
In this article I will talk about three different patterns for writing and maintaining your data migrations:
- Data migrations in
- Data migrations with a set of Rake tasks
- Data migrations with