The acute problem appears to be resolved based on observations over the past hour. We don't have a clear explanation for why the errors began and ended, so further investigation is needed.

The underlying errors that began at roughly 00:00 UTC subsided at roughly 01:00 UTC. We haven't found an explanation for the errors yet, so we are continuing to investigate and monitor the affected systems.