Leading up to the launch of our Public API, we are simplifying some parts of our API which requires technical changes. During one of these changes, a breaking change occurred and interrupted access to Buffer. This impacted all users who were trying to access Buffer on any of our platforms during this time.

This issue is an edge case where service B happened to be deployed before service A, with service B depends on the changes on service A. Because of this, our server crashed and prevented access. The change was quickly reverted to bring users back in.

We are improving our internal documentation around the scenarios where this can occur, along with exploring guardrails to prevent this from happening in future.