As you probably noticed, Visualizer has had a rough couple of days.

What happened?

Yesterday I was hit with the database going down unexpectedly. It took me about 2 hours to get everything up and running, and then I was just happy it all worked.

There were many issues, but one that reemerged and I wanted to address for a while, was running Redis inside the main app container.

I decided to tackle this issue today, and managed to cause even more downtime. Once I succeeded to get separate Redis instances running and remove them from the main image, Fly decided it wasn’t happy about something and just wouldn’t boot the app. Then Docker Hub went down as well, making deploys much harder since it impacted my CI/CD pipeline.

I managed to brute-force it back to life, by nothing other than repeating the same commands over and over again. And then it just started working. Computers, am I right?

After that I decided to go for a quick run to clear my mind, and when I came back I rebuild the entire Docker / Fly setup. With Docker Hub now back online, the deploy pipeline was smooth again, and the new setup feels much more stable¹.

I’m very sorry for all the issues these ~3 hours of combined downtime have caused. This is not how I want this service to run.

A little thank you

To show my appreciation for your understanding and support, everyone will have free Premium this weekend.

Hopefully things will stabilize now. 🤞

P.S. If you’re not aware, you can always check if something is wrong with Visualizer on the Visualizer Status page.