Cause: We experienced an unexpected spike in user requests consuming the allocated resources on the user-facing API interface, resulting in cascading API restarts. Resolution: This has been mitigated by scaling up the resources to prevent future resource exhaustion.