The company Cloudflare recorded a widespread failure on its network this Tuesday morning (18). The problem began around 8:50 am Brasília time and caused unavailability in dozens of popular services around the world. Plataformas like
The company itself confirmed the incident at 8:48 am on its official status page. The advisory mentioned 500 large-scale errors, as well as glitches in the control panel and API. Equipes were mobilized to investigate and fix the problem.
Users reported difficulties on several continents simultaneously. Até the DownDetector website, used to monitor instabilities, was down for a few moments.
Most affected services
X displayed error messages when trying to update the feed. ChatGPT showed blocking alerts related to Cloudflare.
Other impacted services included:
- Amazon and its regional stores
- Discord, with partial interruption of voice and chat
- Canva, preventing project editing
- Gaming sites like League of Legends and Valorant
What does Cloudflare do
The company acts as a protection and optimization layer for more than 24 million domains. Seus services include defense against DDoS attacks, content caching and intelligent traffic routing.
When Cloudflare fails, the original sites remain online but are inaccessible to most users. Isso explains why many internet users had the feeling that “the entire internet went down”.
Incident chronology
The first reports emerged at 8:50 am. Às 9:15 am the company already identified the problem as global.
Updates were published every few minutes. Por Around 10:30 am, traffic began to return to normal in some regions.
At 11:45 am Cloudflare reported that main service was 100% restored. Algumas secondary features were still being verified.
Dependence on digital infrastructure
The episode reinforces the concentration of critical services in a few companies. Semelhante to what happened to AWS in October 2025, the failure demonstrated how a single point can affect millions of users.
Companies of all sizes use Cloudflare to ensure speed and security. The disruption affected everyone from startups to large technology corporations.
Recovery and upcoming updates
The company promised a full report after the analysis was completed. There were no initial signs of a cyber attack.
Global traffic returned to normal throughout the morning. Usuários reported full access to affected services in the early afternoon.
Lessons from the digital blackout
Experts highlight the need for more robust contingency plans. Muitas organizations rely exclusively on a single CDN provider.
The quick identification of the problem allowed correction in less than three hours. Ainda thus, the case serves as a warning for infrastructure diversification.
