TL;DR

AWS’s DNS went dark on October 19, 2025 and took over 24 hours to recover. The newly published Root Cause Analysis reveals a domino effect: a failed DynamoDB DNS update triggered EC2 instance errors, which in turn tripped up Network Load Balancers and knocked out routing.

Quick Dive

For a minute-by-minute breakdown check out the video timestamps (0:00 summary • 5:30 DynamoDB DNS loss • 13:41 EC2 errors • 16:16 NLB errors) or read the full RCA straight from AWS: https://aws.amazon.com/message/101925/

Watch on YouTube