db.utils.OperationalError: connection to server at "pgbouncer.railway.internal" (fd12:dcd8:b376:0:5000:2b:dc88:45bb), port 6432 failed: Connection timed out

Just came back. Second unexplained railway auto-deploy causes downtime this week though ☹

Is there something we can do to know when these auto deploys could happen?

Incredibly sorry about that.

Pretty much right when we solved the incident with the build issues, we experienced an internal network routing issue with a small subset of new hosts we brought online. We have since fixed that and have seen no more 502s or other networking issues.

Appreciate the update, Brody. Was there an update on what happened a week ago as well? Had my redis service autodeploy, and then I had to manually redeploy it again for it to actually work. (I understand if this needs to be answered another, less busy time!)

If it was February 11th, that would be this issue: https://blog.railway.com/p/incident-report-february-11-2026

Despite our best efforts, a small subset of users would have needed to redeploy their workloads manually.

Got it, thanks for the link. Yep, we were definitely one of those impacted!

Hope you have a smooth rest of the week!

You as well, thank you for the kind words!