In the early hours of October 20, 2025, Amazon Web Services (AWS), a cornerstone of global internet infrastructure, experienced a significant outage that disrupted operations for millions of users worldwide. The incident began at 11:49 PM PDT on October 19, when AWS reported elevated error rates across multiple services in its critical US-EAST-1 region. This region, located in Northern Virginia, is one of AWS’s largest and most utilized data centers, handling a substantial portion of internet traffic.
The outage’s impact was widespread, affecting not only AWS’s own services but also numerous high-profile platforms that rely on its infrastructure. Amazon’s e-commerce platform experienced significant disruptions, with customers encountering errors during shopping and checkout processes. Subsidiaries such as Prime Video faced streaming issues, while AWS’s internal support teams struggled to assist customers due to the cascading failures.
The root cause of the outage was identified as DNS resolution problems affecting the regional endpoints for DynamoDB, AWS’s NoSQL database service. DNS, often referred to as the internet’s phonebook, translates human-readable domain names into IP addresses that computers use to identify each other on the network. In this case, the DNS failures prevented applications from locating the necessary server IP addresses, leading to widespread service disruptions.
AWS engineers acted swiftly to address the issue. By 12:26 AM PDT on October 20, they had pinpointed the DNS resolution problems as the root cause. Fixes were implemented, and by 2:24 AM PDT, DynamoDB functionality was restored. However, the outage’s aftermath lingered, impairing a subset of internal subsystems and prompting temporary restrictions on launching new EC2 virtual machines to prevent further instability.
Recovery progressed steadily throughout the morning. By 12:28 PM PDT, most AWS customers and dependent services, including major platforms like Netflix and government sites, reported substantial improvements. Engineers gradually lifted the restrictions on EC2 launches while addressing remaining issues. By 3:01 PM PDT, normal operations were fully restored, enabling smooth operations across the board.
The outage underscored the fragility of DNS in cloud ecosystems and highlighted the need for diversified infrastructure and robust failover mechanisms. AWS emphasized the event’s scope and its rapid response in a detailed post-incident report. While no cyberattack was suspected, the incident served as a reminder of the importance of resilient and redundant systems in maintaining internet stability.
