Amazon's AWS experienced a significant outage due to a major DNS failure linked to a race condition within DynamoDB's infrastructure, affecting users globally for over 14 hours. The incident led to the accidental deletion of all IP addresses for the database service's regional endpoint, causing widespread connectivity issues. In response, Amazon has implemented measures to prevent future occurrences and apologized for the disruption caused to customers.