Between 4:38 PM and 5:35 PM PDT a small number of instances were unavailable due to power loss in a single Availability Zone [usw-az1] in the US-WEST-1 Region. Those customers received notification on their Personal Health Dashboard. During this time, some customers also experienced insufficient-data in the results from EC2 instance status checks for instances that were not impacted by the power event. During this time, CloudWatch alarms may have transitioned into “INSUFFICIENT_DATA” state if set on the delayed metrics. The issue has been resolved and the service is operating normally.
Between 12:40 PM and 1:07 PM PST, we experienced elevated API errors and event delivery latencies in the US-WEST-2 Region. The issue has been resolved and the service is operating normally.