Inside the Storm: How AWS Manages a Major Outage

How AWS deals with a major outage

TLDR: An insider from the AWS Incident Response team provides a detailed account of the recent 15-hour outage in us-east-1. The incident was a complex cascade failure initiated by an unexpected race condition in DynamoDB's internal DNS system, compounded by a separate, simultaneous networking issue.

Link: How AWS deals with a major outage