Originally published at: Cloudflare Outage Takes Down Large Swaths of the Internet - TidBITS
At 6:48 AM on 18 November 2025, Cloudflare posted:
Cloudflare is experiencing an internal service degradation. Some services may be intermittently impacted. We are focused on restoring service. We will update as we are able to remediate. More updates to follow shortly.
What’s not obvious from that dry language is that large parts of the Internet went down as a result, including TidBITS, as Jetpack—a WordPress plug-in that monitors uptime—informed me shortly after. The Cloudflare error page was remarkably helpful, clearly indicating that the problem was in Cloudflare’s network and therefore not something I needed to worry about. That was a big relief for me, since otherwise I would have had to spend potentially hours trying to identify and fix the issue.
Cloudflare said it had resolved all the problems 6 hours later, and CEO Matthew Prince quickly posted a detailed explanation of how the problem was triggered by a database permissions error, causing a file to double in size and crash routing software. He also apologized for the outage:
We are sorry for the impact to our customers and to the Internet in general. Given Cloudflare’s importance in the Internet ecosystem any outage of any of our systems is unacceptable. That there was a period of time where our network was not able to route traffic is deeply painful to every member of our team. We know we let you down today.
For those who aren’t aware, Cloudflare is an Internet infrastructure company that acts as an intermediary between Internet users and about 24 million websites, roughly 15–20% of all websites by some estimates. Cloudflare’s main role is as a content delivery network that caches Web pages in its data centers around the world, making sites load faster for users than if the original servers had to deliver every requested page. The company also protects websites from distributed denial-of-service (DDoS) attacks, provides firewalls to block malicious traffic, helps manage bot traffic, and more. Since basic Cloudflare services are free, it’s extremely popular. (We pay $5 per month for Automatic Platform Optimization to improve Cloudflare’s caching with WordPress.) In the last 30 days, Cloudflare served 1.61 TB of data for tidbits.com, so our server had to deliver only 40 GB—Cloudflare handles nearly 98% of our traffic.
As with the recent Amazon AWS outage in October 2025, this Cloudflare outage underscores that one of the original promises of the Internet’s packet-switching architecture—that it can route around damage—is effectively no longer true. It’s just too compelling for organizations to rely on companies like Amazon and Cloudflare to manage all the hardware, networks, and configurations required to serve terabytes of traffic while defending against nonstop attacks.