We now know why the DNS failed

AWS’s October 19, 2025 outage that knocked DNS offline for over a day boiled down to a hiccup in Dynamo’s replication: it lost its DNS entries and triggered a cascade of control-plane failures. As APIs lagged, EC2 calls started timing out and Network Load Balancers went into a frenzy of health-check failures.

Ten days later, AWS published the full root-cause analysis. The video breaks it down at 0:00 (summary), 5:30 (Dynamo DNS loss), 13:41 (EC2 errors) and 16:16 (NLB errors). For all the gritty details, check out the official RCA: https://aws.amazon.com/message/101925/

Watch on YouTube