A race condition in AWS's DNS management system created an empty record for dynamodb.us-east-1.amazonaws.com, triggering a global cascade. Over 17 million outage reports flooded in from 60+ countries, taking down Snapchat, Robinhood, and even UK tax services. The incident exposed a critical single point of failure in digital infrastructure.

Direct financial losses were estimated at $500-650 million, but the hidden costs were larger: eroded customer trust, stalled innovation, and uncovered insurance gaps. For 15 hours, systems failed because foundational services like EC2's Droplet Workflow Manager depend on DynamoDB for state management, showing how tightly coupled modern cloud architectures are.

This marks the fifth major outage in eight years for AWS's US-EAST-1 region, its oldest and busiest data center hub. While AWS maintains a 99.95% uptime average, these concentrated failures prompt regulators and engineers to rethink multi-AZ deployments, which don't protect against regional infrastructure collapses, and question over-reliance on a single cloud provider.