Amazon Web Services experienced a major data center outage today affecting its US-EAST-1 region — the largest and most heavily used AWS region in the world, located in Northern Virginia. The outage, which began at approximately 10:14 AM EST, caused widespread disruptions to EC2 compute instances, S3 object storage, RDS managed databases, Lambda serverless functions, and dozens of dependent services. Thousands of businesses that rely on US-EAST-1 for their production infrastructure experienced service degradation or complete outages, affecting millions of end users globally.
What Went Down and When
The AWS Service Health Dashboard began showing elevated error rates in US-EAST-1 at 10:14 AM EST today. Within 30 minutes, the impact had cascaded across multiple services. EC2 instances in multiple availability zones within the region became unreachable or experienced severe performance degradation. S3 operations returned intermittent errors, causing failures in applications that depend on object storage for static assets, backups, and data processing. RDS database instances experienced connectivity issues, causing application-level failures for businesses running their databases on AWS managed database services.
Lambda functions, which are used by thousands of companies for serverless computing workloads, experienced elevated error rates and increased latency. API Gateway, which routes traffic to Lambda functions and other backend services, also showed degraded performance. CloudFront, AWS global content delivery network, experienced issues in edge locations that route traffic through US-EAST-1, causing slow load times for websites and applications served through the CDN.
Services Affected
The following AWS services reported issues during the outage: Amazon EC2 (compute instances), Amazon S3 (object storage), Amazon RDS (relational databases), Amazon Lambda (serverless functions), Amazon API Gateway, Amazon CloudFront (CDN), Amazon DynamoDB (NoSQL database), Amazon SQS (message queuing), Amazon SNS (notifications), Amazon ECS and EKS (container services), Amazon ElastiCache (in-memory caching), and Amazon Elastic Load Balancing. The breadth of affected services reflects the deep interdependencies within the AWS service ecosystem — when core infrastructure services like EC2 and S3 experience issues, the impact cascades rapidly through dependent services.
Businesses and Users Impacted
The impact on businesses was immediate and widespread. Several major technology companies that rely heavily on US-EAST-1 reported service disruptions to their customers. Streaming services experienced buffering and playback failures. E-commerce platforms reported checkout failures and slow page loads. Financial services applications experienced transaction processing delays. SaaS companies saw their dashboards and APIs become unavailable. Social media platforms reported issues with content loading and posting.
Downdetector, the website outage tracking service, recorded a massive spike in user reports for dozens of major services simultaneously, confirming the widespread nature of the disruption. Twitter and Reddit were flooded with reports from developers and end users experiencing issues, with the hashtag AWSDown trending globally within an hour of the outage beginning.
AWS Response and Timeline
AWS engineers acknowledged the issue on the Service Health Dashboard at 10:47 AM EST, approximately 33 minutes after the first signs of degradation appeared. The initial acknowledgment described increased error rates and latency affecting multiple services in US-EAST-1 without providing details about the root cause. Subsequent updates were posted every 30-45 minutes as engineers worked to identify and resolve the underlying issue.
By 1:30 PM EST, AWS reported that the root cause had been identified and that recovery actions were underway. The company indicated that the issue was related to a networking problem within the US-EAST-1 data center infrastructure, though full technical details were not immediately disclosed. Services began recovering progressively, with some returning to normal operation while others continued to show elevated error rates.
Full service restoration was declared at approximately 3:45 PM EST, approximately 5.5 hours after the outage began. AWS committed to publishing a detailed post-incident report within 72 hours explaining the root cause, the timeline of events, and the steps being taken to prevent similar incidents in the future.
Financial Impact
The financial impact of the outage is difficult to quantify precisely, but industry analysts estimate that a 5-hour outage affecting US-EAST-1 could cost AWS customers hundreds of millions of dollars in lost revenue, productivity, and recovery costs. E-commerce companies lose an estimated 2.5 million dollars per minute of downtime during peak hours. Financial services firms face regulatory penalties for service availability failures in addition to direct revenue losses. SaaS companies face customer churn and SLA penalty payments when their services go down due to cloud provider outages.
AWS itself faces financial consequences through SLA credits — the company's service level agreements guarantee 99.99% monthly uptime for most services, and a 5.5-hour outage significantly exceeds the allowable downtime under these agreements. Affected customers are entitled to service credits, though the credit amounts are typically a fraction of the actual business impact experienced.
Why US-EAST-1 Outages Are So Impactful
US-EAST-1 is the oldest and largest AWS region, hosting a disproportionate share of AWS global infrastructure and customer workloads. Many AWS services were first launched in US-EAST-1 and have their primary infrastructure there. The region also hosts significant AWS internal infrastructure, meaning that issues in US-EAST-1 can affect AWS management plane operations globally, not just customer workloads in that region.
The concentration of workloads in US-EAST-1 reflects a historical pattern where early AWS adopters chose the region for its low latency to the US East Coast and its status as the most feature-complete region. Despite AWS strong recommendations to architect for multi-region resilience, many organizations continue to run single-region deployments in US-EAST-1, making them vulnerable to exactly this type of regional outage.
Lessons for Cloud Architecture
Today outage is a reminder of the importance of multi-region architecture for business-critical applications. Organizations that had implemented active-active or active-passive multi-region deployments were able to fail over to alternative regions and maintain service continuity during the outage. Those running single-region deployments in US-EAST-1 had no such option and experienced the full impact of the disruption.
AWS recommends that customers architect their applications to be resilient to availability zone failures at minimum, and to regional failures for the most critical workloads. Services like Amazon Route 53 health checks, AWS Global Accelerator, and CloudFront can be used to automatically route traffic away from a degraded region. While multi-region architecture adds complexity and cost, today outage demonstrates that for many businesses, the cost of downtime far exceeds the cost of resilience.
AWS History of Major Outages
Today outage is not the first time US-EAST-1 has experienced significant disruptions. The region has been the epicenter of several major AWS outages over the years, including a December 2021 outage that took down large portions of the internet for several hours, a November 2020 outage that affected Kinesis and cascaded to dozens of dependent services, and multiple smaller incidents affecting individual services. Each outage has been followed by detailed post-incident reports and commitments to improve resilience, yet the fundamental challenge of operating infrastructure at AWS scale means that outages, while rare, remain an inevitable reality of cloud computing.
