NEW YORK – A widespread disruption to Amazon Web Services’ (AWS) CloudFront content delivery network (CDN) on February 10, 2026, impacted numerous online services, highlighting the critical role CDNs play in modern internet infrastructure and raising questions about redundancy and resilience. The outage, which began around 9:15 PM UTC, saw CloudFront returning DNS resolution failures – effectively making websites and applications inaccessible to users. While the core DNS issue was resolved within approximately an hour, full recovery, including the propagation of changes to new distributions and TLS certificate provisioning, extended until around 4:00 AM UTC on February 11.
The incident wasn’t isolated. According to reports from IsDown.app, a monitoring service, users began experiencing problems 23 minutes before AWS issued its initial status update, demonstrating the value of crowd-sourced monitoring in early incident detection. The cascading effect of the CloudFront failure impacted at least eight interconnected AWS services, including Route 53, API Gateway, WAF, AppSync, Pinpoint, Transfer Family and VPC Lattice. Downstream, the disruption rippled through major platforms like Salesforce, Adobe, Discord, and Claude, affecting millions of users globally.
What Happened: A Cascade of DNS Failures
The root cause, as identified by AWS, was a DNS resolution failure within CloudFront’s infrastructure. This meant that when users attempted to access content delivered through CloudFront, the Domain Name System (DNS) – the internet’s phonebook – was unable to translate the domain name into the correct IP address, resulting in “NXDOMAIN” responses, indicating that the domain simply didn’t exist. This isn’t a simple website being down; it’s a fundamental failure in how the internet directs traffic.
The failure quickly spread due to the interconnected nature of AWS services. Route 53, AWS’s scalable DNS service, was likewise affected, exacerbating the problem. API Gateway experienced increased error rates and latencies, while AWS WAF, a web application firewall, reported errors due to its dependency on CloudFront. Even services seemingly less directly connected, like Amazon Pinpoint (a marketing communication service) and Amazon Transfer Family (managed file transfer), experienced increased errors.
CloudFront and the Importance of CDNs
Amazon CloudFront is a globally distributed network of servers designed to cache content closer to complete-users, reducing latency and improving website performance. As a content delivery network, it’s a crucial component of the modern internet, handling a significant portion of web traffic. The service delivers both static content – images, videos, CSS files – and dynamic content, ensuring a fast and reliable user experience. A distribution, as defined by AWS CloudFormation, essentially tells CloudFront where to deliver content from and how to manage that delivery.
Although, this incident underscores the inherent risk of relying on a single CDN provider. While CDNs offer significant benefits, a widespread outage like this demonstrates the potential for a single point of failure to disrupt a vast ecosystem of online services. The reliance on CDNs has grown exponentially in recent years, making their stability paramount.
Troubleshooting and Mitigation
AWS has provided documentation outlining steps to troubleshoot and prevent similar errors in the future. The documentation highlights the importance of understanding how CloudFront Functions interact with origin mutual TLS (mTLS). CloudFront Functions, a lightweight serverless compute service, can be used to customize content delivery, but specific limitations apply when used with mTLS. For example, while you can switch between origin mTLS origins using the same client certificate, or conditionally disable mTLS, certain modifications are restricted.
For those providing content through CloudFront, the documentation suggests reviewing configurations and ensuring proper setup of mTLS, if utilized. The incident also highlights the necessitate for robust monitoring and alerting systems to detect and respond to outages quickly. The fact that IsDown.app users reported issues before AWS’s official update underscores this point.
What’s Next?
AWS has not yet released a detailed post-mortem analysis of the February 10th outage, but it is expected to provide a comprehensive report outlining the root cause, the steps taken to resolve the issue, and the measures being implemented to prevent recurrence. The incident is likely to prompt a broader industry discussion about CDN redundancy and the need for multi-CDN strategies, where organizations utilize multiple CDN providers to mitigate the risk of a single point of failure.
The next scheduled AWS re:Invent conference, set to begin November 30, 2026, in Las Vegas, will likely feature sessions addressing CDN reliability and best practices for building resilient applications.
This event serves as a stark reminder of the fragility of the internet infrastructure and the importance of proactive measures to ensure its stability. As more and more services rely on CDNs, the need for robust monitoring, redundancy, and incident response capabilities will only continue to grow.
