Amazon Web Services (AWS) is showing signs of recovery after a major outage disrupted dozens of popular apps and websites early Monday. Services affected included Snapchat, Roblox, Fortnite, Canva, Duolingo, Zoom, Amazon, MyFitnessPal, and Signal.
The disruption began around 3:11 a.m. ET and was caused by high error rates and latency in AWS’s DynamoDB service, combined with DNS resolution issues in the US-East-1 region. AWS confirmed the problem on its status dashboard, noting impacts on over 20 services, including CloudFront, AWS Config, and the Security Token Service.
Monitoring site Downdetector recorded more than 2,000 reports in the United States, with additional spikes in the UK and Europe. Banking apps such as Lloyds and Halifax, government services including HM Revenue & Customs (HMRC), and telecom providers like Verizon and Vodafone were also affected.
Gaming platforms were hit particularly hard, with millions of Fortnite and Roblox players unable to log in. Canva users encountered blank screens mid-design, Duolingo lessons stalled, and Zoom calls faced connectivity problems.
Amazon checkout errors occurred, while smart home devices like Ring doorbells and Alexa went offline. Streaming platforms including Prime Video, Hulu, and Disney+ reported interruptions, and cryptocurrency exchanges such as Coinbase and Robinhood experienced intermittent issues.
By midday UTC, AWS reported “significant signs of recovery.” Engineers identified DNS resolution as the main cause and implemented fixes, leading to a drop in error rates. While services such as Snapchat and Roblox showed improvement, some Zoom and Amazon features remained unstable. AWS continues to monitor the situation, with a full resolution expected in the coming hours.
The incident highlights the risks of relying on centralized cloud infrastructure. AWS supports over a third of Fortune 100 companies, and previous outages, including a 2021 event affecting Netflix and Disney+, have prompted calls for diversified hosting strategies.
Users are advised to check service-specific status pages and retry access as recovery continues.
