A failure in Cloudflare’ s infrastructure has generated this Tuesday one of the most important interruptions in Internet services in recent months. Popular sites such as X (formerly Twitter), ChatGPT, bet365 or Letterboxd stopped working, displaying error messages directly linked to the network services company.
Cloudflare confirmed the issue with an initial statement, “Cloudflare is aware of an issue that could affect a number of customers and is investigating.” The company reported that it was continuing to investigate the situation and would update information as it progressed.
The impact was so widespread that even Down Detector, the site that monitors crashes, was affected. When it managed to load, it showed a massive spike in reported incidents, especially on platforms such as OpenAI, bet365 or League of Legends.
Error and crash messages on dozens of platforms
The crash triggered multiple error messages. Many users saw the notice that there was an “internal server error on the Cloudflare network” and that they should “try again in a few minutes.”.
In the case of bet365, the message was particularly alarming for users: “Sorry, you have been blocked. You are unable to access bet365.com” Although, as The Independent explained, there was no cause for concern. The situation was simply due to Cloudflare not working properly.
The same happened with ChatGPT, which displayed a generalized warning: “Please unblock challenges.cloudflare.com to continue” Again, without the user needing to do anything more than wait for the service to be restored.
Partial recovery and ongoing investigations
Throughout the day, Cloudflare issued several updates. In one, it explained, “We are seeing a recovery of services, but customers may continue to experience higher than normal error rates as we continue remediation efforts.”
However, shortly afterwards he added that the problem was not completely resolved: “We continue to investigate this issue”. This communication came after detecting new failures, even after a brief period in which some websites started working.
The company also posted a more detailed notice on its status page: “Cloudflare is aware of an issue affecting several customers and is investigating: widespread 500 errors, and Cloudflare’s control panel and API are also failing” It further indicated that it was working to understand the full scope of the bug and mitigate it.
A massive failure comparable to the AWS crash
This incident is reminiscent of what happened just a month ago, when a technical problem affected Amazon Web Services (AWS) services for hours. Like then, Cloudflare’s outage was so widespread because its technology supports thousands of sites, usually hiding behind the scenes.
Although service began to recover at some sites, the company never confirmed that the cause had been fully identified or that the issue had been resolved.
How the problem started and what is known so far
The first failures were detected around 11:30 am (UK). In the initial minutes, some pages could partially load if refreshed, indicating an intermittent problem. Cloudflare posted its first warning about 15 minutes later and continued to report without providing details on the cause.
According to the latest media reports, the situation was still under study and errors persisted in some services. Several platforms (including X and others with millions of users) began to restore normality progressively, although with residual failures.
Pending a definitive technical explanation from Cloudflare, the incident confirms once again the global dependence on large infrastructure providers for the functioning of the digital ecosystem.
