4 posts tagged with "ci"

Hi everyone!

I'm very happy to share with you some of the recent improvements to Atlas, specifcially around GitHub Actions. In August of last year, we released our first version of the GitHub Actions experience for Atlas. It was a modest start, which included the ability to verify the safety and correctness of schema migrations during the CI process.

Over the past year, we have slowly added more features to the GitHub Actions experience, including the ability to sync migration directories to Atlas Cloud, deploy migrations, and even install Atlas. As often happens with quickly evolving systems, we felt that the API became complex, carrying over use cases and experiences that have become obsolete or superseded by better ones since the initial release.

At Ariga, the team developing Atlas, we have written a document named the "R&D Manifesto", which lists some the principles that we commit to as individuals and as an organization. One of them is "Obsess over APIs and DevEx" - we believe that the key to building a successful product is to provide the best possible experience to our users, and that starts with clear, consistent and composable APIs that empower our users to achieve amazing feats of engineering.

With that in mind, our team has been working hard in the past few weeks to revamp the GitHub Actions experience for Atlas. Here's a quick summary of the changes:

1. We've moved all actions into a single repo - ariga/atlas-action. (With the exception of ariga/setup-atlas.)
2. The API has been reviewed and updated to make sure it is consistent among the different actions and with the rest of the Atlas ecosystem.
3. We've rewritten the code in Go, which is the language we use for all of our internal tools. This allows us to share code between the CLI and the GitHub Actions, and to provide a more consistent experience between the two. In addition, looking forward we have greatly simplified the process of adding new GitHub Actions as needed.

As applications evolve, database schema changes become inevitable. However, some types of changes can break the contract between the backend server and the database, leading to errors during deployment and application downtime. Ensuring that these breaking changes are detected and prevented is crucial for maintaining system stability and delivering a seamless user experience. In this post, we will explore how Atlas can help you prevent breaking schema changes from impacting your application.

Losing data is painful for almost all organizations. This is one of the reasons teams are very cautious when it comes to making changes to their databases. In fact, many teams set explicit policies on what kinds of changes to the database are allowed, often completely prohibiting any change that is destructive.

Destructive changes are changes to a database schema that result in loss of data. For instance, consider a statement such as:

ALTER TABLE `users` DROP COLUMN `email_address`;

This statement is considered destructive because whatever data is stored in the email_address column will be deleted from disk, with no way to recover it.

With the release of v0.5.0, we are happy to announce a very significant milestone for the project. While this version includes some cool features (such as multi-file schemas) and a swath of incremental improvements and bugfixes, there is one feature that we're particularly excited about and want to share with you in this post.