Overview
- The incident began early on October 20 with elevated error rates and latency across multiple AWS services, with Downdetector logging more than four million user reports worldwide.
- AWS reported full regional recovery by about 19:01 Brasília time on Monday, noting that some workloads faced lingering delays as queues and backlogs were processed.
- The disruption affected a broad range of platforms, including Snapchat, Fortnite, Roblox, Canva, Prime Video, Alexa, Coinbase, Robinhood, and Lyft, according to company updates and monitoring sites.
- Technical updates described a DNS-related failure impacting DynamoDB in US‑EAST‑1, which cascaded into issues for EC2, Lambda, CloudWatch, and Network Load Balancer health checks before mitigations took hold.
- In Brazil, users reported problems with services such as Pix and certain payment apps, though direct links to the AWS incident were not confirmed in all cases.