A global DNS issue may have resulted in the Power Outage Systems being partially or entirely inaccessible, from 21:30 UTC to about 22:20UTC. This was a global issue effecting all Azure customers, a huge section of the World Wide Web.
This also caused the Power Outage Systems to stop updating power outage data. During this event a large number of utility websites where also offline, triggering a failsafe mechanism designed specifically for these situations. Preventing the Power Outage System from being updated with bad data (in this case zero outage counts for the utilities also effected by the global DNS outage.)