A major outage that hit Cloudflare’s global network has led to widespread internet outages, affecting popular platforms like Elon Musk’s X, ChatGPT, and even Downdetector, the number one place people go to understand internet outages.
Cloudflare, the world’s leading web infrastructure and security company, confirmed that it experienced a significant issue with its global network in the early hours of Tuesday morning, causing widespread outages across the internet. The issue affected a variety of services, including social media platform X, OpenAI’s ChatGPT, and outage tracking site Downdetector, among many other sites and apps that are experiencing issues. Even some news sites, including Axios, appear to be completely unavailable at the time of writing due to the Cloudflare outage — visitors are being asked to “unblock” a Cloudflare address in order to continue.
According to Cloudflare’s system status page, the company is currently investigating the issue, which is causing HTTP 500 errors affecting Cloudflare’s dashboard and API. These server-level errors indicate that the problem lies with Cloudflare’s infrastructure, not individual websites or servers. The company claims in one of its most recent updates that “The issue has been identified and a fix is being implemented.”
As a result of the outage, X has been experiencing intermittent issues, with availability being patchy over the past hour. Downdetector, which ironically relies on Cloudflare’s services, was also affected by the outage. The site saw a peak of 11,201 problem reports at 6:37 a.m. ET, with the number of reports gradually decreasing to 6,570 at the last data point. This decrease could suggest that the fix is slowly spreading across the internet, or that fewer people are reporting the issue as it becomes more widely known.
Cloudflare underwent scheduled maintenance at 5:00 a.m. ET, according to the company’s website. This suggests that the outage may have been the result of a failed internal change, rather than a cyberattack or other external issue.
In an attempt to mitigate the issue, Cloudflare has disabled access to WARP in London. WARP is a tool designed to secure internet connections by encrypting all traffic from a user’s device, primarily used as a privacy tool for consumers and corporate clients. Unlike a VPN, WARP does not hide a user’s IP address; it simply encrypts their traffic.
The outage also affected the 3D printing community, with popular platforms like Printables and Thangs displaying HTTP 500 error codes when users tried to access them.
Breitbart News reported in October that a massive outage at Amazon Web Services affected much of the internet:
The root cause of the outage was traced to a technical update error in the DynamoDB API, a key cloud database service that stores user information and other critical data for numerous online platforms. The update issue affected the Domain Name System (DNS), which helps applications locate the correct server addresses. As a result, applications were unable to find the IP address for the DynamoDB API and were unable to establish a connection.
DNS is one of the core technologies of the internet. It is often compared to a “phone book” for Internet servers, translating the addresses we type in like amazon.com into IP addresses that are used to connect us to the appropriate server. If DNS were the phone book, then a bug in AWS’s DynamoDB would be a phone book listing the wrong number for every person and business—resulting in a lot of failed connections.
Outages from major providers like Cloudflare and Amazon Web Services show that the Internet is not the robust and redundant online service it is made out to be, due to its reliance on a few huge companies to keep data flowing.