Article

On February 20, 2026, Cloudflare's global network suffered a 90-minute outage affecting 15% of its capacity, triggered by a misconfigured rule in its Traffic Manager system. The incident started at 14:51 UTC due to an engineer's error during a routine update, which incorrectly deprioritized multiple data centers.

The outage led to billions of failed requests, primarily impacting users in Western and Central Europe. In response, Cloudflare is mandating peer reviews for rule changes, adding a blast radius preview tool, and restructuring data center naming conventions. The incident highlights systemic risks in concentrated internet infrastructure, prompting discussions about operational safeguards.