Incident started around 15:51 UTC, resolved around 18:11 UTC. After investigation, we found that no customer queries were affected; only our internal metrics systems were impacted.

We are investigating reduced query throughput for some data source plugins in multiple regions. Some queries may fail or be slow.