Overview
- AWS reported increased errors at about 3:11 a.m. ET on Oct. 20 and said the DNS/DynamoDB problem was fully mitigated by roughly 6:35 a.m., though elevated API errors and connectivity issues persisted for hours.
- AWS applied mitigations including throttling and rate‑limiting new EC2 instance launches in US‑EAST‑1 to stabilize systems and continues clearing request backlogs, leading to intermittent delays for some customers.
- Downdetector recorded about 6.5 million user reports with more than 1,000 sites affected, spanning gaming, streaming, finance, airlines, government services and IoT platforms.
- Companies including Coinbase, Perplexity and Signal confirmed disruptions tied to AWS as users reported issues with services such as Snapchat, Fortnite, Roblox, HMRC, United Airlines and Amazon’s own Alexa and Prime Video.
- AWS said investigations are ongoing, citing an internal EC2 network and a subsystem monitoring network load balancers as contributing factors, and it plans to publish a post‑incident summary.