We now know why the DNS failed: on October 19, 2025 AWS rolled back a DynamoDB DNS record in a way that poisoned recursive resolvers for over 24 hours, sparking a massive outage. This DNS hiccup cascaded into EC2 metadata request failures and, eventually, Network Load Balancer errors, dragging key AWS services offline.
Ten days later, AWS’s root cause analysis lays out the blow-by-blow: the bad DNS change at 5:30 kicked off resolver failures, EC2 errors piled up by 13:41, and NLB health checks collapsed at 16:16. It’s a stark reminder that even a tiny DNS misstep can bring a cloud giant to its knees.
Watch on YouTube
Top comments (0)