At 11:10 PM PST on December 6 a small number of underlying hosts experienced power loss, affecting some customers in sae1-az3 in the SA-EAST-1 Region. Impacted customers were automatically notified via the Personal Health Dashboard. We quickly identified root cause and began working toward recovery. At 2:25 AM PST on December 7 additional hosts were impacted. Shortly after that we began to successfully mitigate impact and as of 4:00 AM PST we have begun to see recovery. We continue to work toward full resolution.
Between 5:25 PM and 7:52 PM PDT, some AWS Batch Compute Environments transitioned to INVALID in the US-WEST-2 Region. The issue has been resolved and the service is working normally.