The issue has been resolved. We apologize for the downtime.

This morning, our compute cluster experienced issues reaching internal APIs. This is connected to an issue at our hosting provider (https://status.scaleway.com/incidents/11x4s3y16p9g).

This issue was eventually resolved by Scaleway staff, but it did create a scheduling problem that prevented many Redis instances from starting, thereby causing downtime for some applications (including Mastodon).

We are currently taking steps to mitigate this issue.