TL;DR

We finally know what brought AWS to its knees on October 19, 2025: a rare DynamoDB DNS replication bug that knocked out name resolution worldwide. Once DNS went haywire, EC2 instances failed to register and Network Load Balancers couldn’t route traffic, spiraling into a full-blown, 24-hour outage.

AWS’s official RCA walks through each domino—DNS, EC2 errors and NLB issues—to show how a tiny glitch can cascade into a massive service disruption. Check out the full report for all the gritty details.

Watch on YouTube