AWS October 19, 2025 Outage: The TL;DR

AWS went dark for over 24 hours on October 19, 2025, and only shared the full root-cause analysis 10 days later. It all started when Dynamo’s DNS configuration went haywire, knocking out name resolution across key services.

On top of that initial DNS failure, EC2 instances started throwing errors and Network Load Balancers couldn’t route traffic properly, creating a cascading meltdown that took more than a day to fully recover. The detailed RCA breaks down exactly how each layer tripped over the one below it.

Watch on YouTube