One of our DB clusters suffered a peak of processing load due to an internal management operation causing a temporary outage to the clients relying on the affected cluster.
We are quickly working on a new access policy to control these kind of operations.