This article explains how Authress maintained service availability despite the significant AWS outage on October 20th. It discusses the importance of reliability in their authentication services and the architectural strategies they implemented to achieve a five-nines SLA.

The article discusses the inevitability of outages and the hidden dependencies in business architectures that rely on cloud services. It emphasizes the need for robust backup plans and testing strategies, like brownouts and Chaos Monkey, to prepare for potential failures. The author argues that businesses must recognize and address these risks to avoid being blindsided by downtime.