by javhd on 2025 Nov, 20

On November 18, 2025, Cloudflare — one of the world’s largest internet infrastructure and security providers — experienced a major global outage that disrupted access to numerous high-traffic platforms, including X (formerly Twitter), ChatGPT, Canva, and many other services. The incident, which lasted several hours, caused widespread connection errors and slowdowns across a significant portion of the internet.

According to Cloudflare, the outage began around 11:20 UTC and was triggered by an internal configuration issue rather than a cyberattack. A change involving database permissions caused a bot-management feature file to generate far more entries than expected, exceeding system limits and crashing a core service responsible for handling global traffic. The company emphasized that no malicious activity was detected, confirming that the problem was entirely self-inflicted.

As the failure spread, users worldwide encountered HTTP 5xx errors, inaccessible websites, and prolonged loading times. Because Cloudflare sits in front of millions of domains, the disruption had a cascading effect across various industries, from social media and AI tools to e-commerce and enterprise services.

By 14:30 UTC, Cloudflare managed to restore a majority of traffic routing functionality, and by 17:06 UTC, the company announced that all systems had returned to normal operation. Engineers continued monitoring and verifying components to ensure long-term stability.

The incident reignited concerns about the global dependency on a small number of internet “gatekeepers.” Experts noted that large-scale outages highlight how fragile the modern internet ecosystem can be when a single provider experiences technical failure.

Cloudflare issued a formal apology and stated that a full internal review is underway. The company promised to strengthen safeguards, adjust system limits, and improve validation checks to prevent similar outages in the future.