Overview
- The incident began around 3:11 a.m. ET on Oct. 20 in US‑EAST‑1, causing elevated error rates and latency across dependent services.
- Amazon reported DNS resolution failures impacting the DynamoDB API and internal EC2 networking and load balancers, cascading into outages for roughly 64–71 AWS services.
- Downdetector recorded more than 6.5 million user reports as platforms from Snapchat, Canva and Fortnite to Coinbase, Robinhood and Lloyds Bank experienced disruptions.
- Financial services and payment rails in several countries saw processing errors, and AWS said it had no evidence of data loss.
- By Oct. 21 operations largely returned to normal with some backlogs clearing over hours, and AWS pledged a detailed post‑incident report as customers weigh redundancy and multi‑cloud strategies.