Breaking: AWS Outage Takes Down Major Sites Worldwide—Service Status, Impact, and Recovery Timeline

Updated January 22 2026 – 16:00 UTC The Amazon Web Services (AWS) cloud platform is experiencing a major outage centered on its critical US-EAST-1 region, disrupting thousands of websites and apps worldwide. According to the AWS Health Dashboard, engineers confirmed “network connectivity issues” beginning at 07:29 PDT/15:29 UTC and are “actively working toward recovery” while traffic is being rerouted to healthy infrastructure. Among the brands reporting downtime are Fortnite, Snapchat, Duolingo, Perplexity AI, Airtable, Canva, and Ring, as well as banking portals and smart-home services. Internal Amazon properties—including portions of Amazon.com and Alexa—also show intermittent errors. Why US-EAST-1 outages hurt so much The US-EAST-1 (Northern Virginia) cluster handles a disproportionate share of DNS, authentication, and control-plane traffic for AWS. When this hub stumbles, dependent regions often feel the shockwave, turning a single-region fault into a global event. Analysts note that today’s incident once again highlights architectural single points of failure despite AWS’s multi-AZ design philosophy. Timeline of today’s AWS outage • 07:29 PDT – AWS detects elevated network errors in US-EAST-1. • 07:45 PDT – Major consumer apps begin reporting “503 Service Unavailable” and login failures. • 08:30 PDT – AWS starts traffic engineering to alternative paths; partial recovery for S3 and DynamoDB. • 10:15 PDT – Outage peaks with more than 20,000 public incident reports on DownDetector. • 15:00 PDT – AWS says it is “seeing steady improvement,” but pockets of latency persist. Services most affected 1. Content delivery: CloudFront, API Gateway 2. Data stores: DynamoDB, Aurora, Redshift 3. Compute: Lambda, Fargate, EC2 Auto Scaling 4. Communication: SNS, Chime SDK 5. Identity: Cognito, IAM token issuance How to check AWS status and mitigate impact • Monitor the AWS Health Dashboard and RSS feeds for per-service recovery messages. • Implement multi-region failover via Route 53 or third-party DNS to avoid hard dependency on US-EAST-1. • Cache critical configuration data locally so your application can operate in “degraded but functional” mode. • For SaaS teams, update status pages and employ circuit-breakers to keep user interfaces responsive even when backend calls fail. What happens next AWS historically delivers post-mortem root-cause analyses within 10 business days, detailing any hardware failures or software regressions involved. Enterprises should review today’s events as a stress test of their own resilience plans and consider adding cross-region replication for both data and CI/CD pipelines. The bottom line With roughly one-third of the internet touching AWS in some fashion, an AWS outage is effectively a partial internet outage. Until full remediation is confirmed, expect intermittent errors when streaming media, playing online games, or accessing finance and productivity tools tied to US-EAST-1.