AWS might have had back-to-work traffic in lots of domains simultaneously.

Or maybe their monitoring and response staff was just coming back online.