Seq on Railway

Posted Jun 9, 2024

I’m using Railway for some side projects. It’s like Netlify for traditional, non-serverless web apps, and it works really well: I can just push to git and the code is built and deployed, without having to worry about hosting infrastructure. It even supports structured logging by reading JSON-formatted logs directly from stdout.

But Railway’s built-in log explorer is pretty rudimentary, so you might want something more advanced if you’re dealing with a lot of logs. At work, we use Seq for application logging, and it too works really well: I can query structured logs with a flexible query language, set up dashboards and alerts, plot occurrences over time, and so on.

But until now, deploying Seq on Railway has been a bit of a pain. While Seq has an official Docker image, it needs a little configuration to make it work with Railway’s port binding and private networking features. My new template should make this process easier:

It contains a single service that spins up the Seq image from Docker Hub, but with some preconfigured options that should make it just work, as any Railway user would expect 😉