Amazon Reveals Technical Fault Behind Widescale AWS Service Outage

Amazon Web Services experienced a major outage that affected millions of customers and Amazon’s own operations on 19 and 20 October 2025, with the company later confirming a DNS resolution issue at regional DynamoDB endpoints as the root cause. According to aboutamazon[.]com, the disruption lasted about two hours and thirty-five minutes, beginning at 11:49 PM PDT on 19 October and ending at 2:24 AM PDT on 20 October, with the US-EAST-1 region bearing significantly increased error rates.

AWS engineers identified the DNS problem at 12:26 AM PDT and began mitigation, resolving the core DynamoDB DNS issue by 2:24 AM PDT, though a small subset of internal subsystems remained impaired afterwards. To aid recovery, AWS throttled certain operations, notably new EC2 instance launches, for a period before gradually lifting the throttling as services and customer systems recovered.

By 3:01 PM PDT on 20 October, AWS announced all services had returned to normal, and the company continues to monitor health while publishing a post-event summary outlining changes to prevent similar incidents.