On 18 November 2025, anyone trying to open Twitter, sorry X, got a very different kind of feed: blank timelines, error pages and apps that just refused to refresh. It looked at first like “X is down again,” but this time the issue was bigger than one platform. The problem traced back to Cloudflare, one of the internet’s main infrastructure providers, which ran into a major outage and took a lot of high traffic sites down with it for a while.
To understand what happened, it helps to know what Cloudflare actually does. Think of it as the traffic and security layer that sits between you and many of the sites you use every day. It speeds things up with caching, protects against attacks and handles DNS and routing. When that middle layer has a bad day, the websites behind it suddenly look broken, even if their own servers are fine. That is why X could be “down” even though its own systems were not the original source of the failure.
During the outage, users across multiple countries reported that X would not load, timelines would not refresh and posts, replies or DMs would fail to send. On desktop, some people saw generic 5xx style errors. On mobile, the app often sat on a loading spinner forever. For creators, news accounts and brands that rely on X for real time communication, it felt like someone had yanked the power cable out of the global conversation.
Of course, X was not alone. Whenever Cloudflare has trouble, many other tools stumble too. Various AI tools, SaaS dashboards, ecommerce sites and content platforms all showed intermittent errors or became completely unreachable. Outage trackers like Downdetector lit up, which is usually the first sign that an infrastructure layer rather than a single app is in trouble. For a while, the most reliable status page on the internet was people complaining in group chats.
Cloudflare engineers worked on rerouting and stabilising their network, and services began to flicker back to life as routes recovered. Some users could access X in one region while others still saw errors, simply because DNS and network changes take time to propagate globally. Eventually, performance normalised, timelines loaded and life returned to its usual mix of memes, arguments and breaking news.
The bigger story here is about how dependent the modern internet has become on a handful of infrastructure providers. When a CDN or DNS giant like Cloudflare sneezes, huge parts of the web catch a cold. For everyday users, that is mildly annoying. For brands, publishers and creators who put all their audience eggs in one basket, it is a reminder to diversify. Email lists, multiple social channels and owned websites are still very boring, very good insurance for days when a single outage makes a “global town square” suddenly go dark.