Cloudflare, a company that invisibly holds up much of the internet, has become a lot better known this year.
It’s one of the many huge web infrastructure providers that allow websites to serve their pages to users. These businesses mostly remain in the background until something goes wrong.
That happened a few weeks ago, and again on Friday morning, when many of the world’s biggest websites stopped loading as normal.
In November, visitors to X, formerly known as Twitter, and other sites instead saw a message from Cloudflare itself. It told them that there had been an error, that the company’s systems were not working properly, and that they should “try again in a few minutes”.
Cloudflare describes itself as “one of the world’s largest networks”, with “millions of internet properties”. “Today, businesses, non-profits, bloggers, and anyone with an Internet presence boast faster, more secure websites and apps thanks to Cloudflare,” its website says.
It does that through a whole array of products. But its most famous and central offering is a set of technologies that ensure websites can stay online when they are receiving high amounts of traffic – whether that is coming from an unusually high number of visitors, or an attack designed to take the website down.
In the simplest form of the internet, a computer asks for a website and is provided with the data by a server. This is then loaded onto the screen. But in that simple setup, the server can become overwhelmed by requests, and could run slowly or break entirely as a result.
Cloudflare and similar companies sit between the computers and the websites themselves. These internet infrastructure providers can then use their more resilient data centres to provide website data quickly and reliably.
That means, however, that problems at companies such as Cloudflare can take down websites quickly. And because their tools are used by such an array of different websites, those problems can quickly knock down a variety of seemingly unconnected pages. That is what happened, too, when another web infrastructure provider – Amazon Web Services, or AWS – suffered a similar outage.
The companies involved normally operate unseen, because they are simply providing the websites that people want to visit – but they suddenly become more conspicuous when they stop working.
