Resolved -
From 20:03 - 20:35 UTC, some API endpoints incorrectly returned HTTP 4xx error codes. This was caused by a bad code change that resulted in feature flags not being properly applied. The bad code change was rolled back at 20:30 UTC and full service was restored by 20:35 UTC.
Jan 16, 20:00 UTC
Resolved -
From 22:03 - 22:26 UTC, requests to tasks deployed on the corrupted container instance incorrectly failed with HTTP 401 status codes. The corrupted instance was terminated at 22:26 at which point service was restored. The container instance was corrupted due to a race condition triggered by how AWS ECS launches tasks on container instances. We have since made changes to safeguard against this race condition.
Jan 15, 22:00 UTC