How to implement the Outbox pattern in Go and Postgres

How and why to use the Outbox pattern to build a robust event-driven system.

I was at a ContainerDays conference recently and attended a great talk from Nikolay Kuznetsov about the Outbox pattern and resilient system design. It seemed like a powerful solution to a common problem in distributed systems, so I decided to dive deeper, and I want to share what I've learned and summarize for our readers in this post.

The challenge in event-driven systems

In modern, event-driven architectures, services often communicate asynchronously using a message broker. A typical flow looks like this: a service receives a request, updates its own database, and then publishes an event to notify other services about the change. Or these two actions happen in parallel.

Here's the problem: what happens if the database commit succeeds, but the subsequent call to the message broker fails? Maybe the broker is temporarily down, or there's a network glitch. Or what if the database is not available at that time? Or what if the program somehow crashes?

You end up in an inconsistent state. Your local database has the new data, but the rest of the system never gets the notification. This is a serious issue because the database operation and the message publishing are not atomic - they don't succeed or fail as a single, indivisible unit.