Cloudflare experienced a global outage on Nov. 18, 2025, that disrupted access to several major websites and online services. The incident began when the company pushed a configuration file that was larger than expected, which caused key systems to fail and led to widespread errors across its network.
According to Cloudflare, the issue started around 11:20 UTC, when internal systems that use the file for traffic management began to crash. The file, which is part of the company’s Bot Management tool, had grown significantly after a permission change allowed it to expand beyond its normal size. Systems that depended on the file were not designed to process it at that scale, causing a chain reaction that affected global traffic.
As the file spread across different data centers, users began to see connection failures, slow loading times, and unavailable services on websites that rely on Cloudflare. Platforms such as social networks, messaging services, and content sites were among those affected.
Cloudflare engineers halted the rollout once they identified the source of the problem. They replaced the faulty file with an older version that had already been validated and began restoring services. Most functions were operational again by early afternoon UTC, and full recovery followed later in the day.
The company apologized for the disruption and outlined steps to prevent similar incidents. These include adding stronger checks for configuration files, tightening controls on internal permissions, and improving safeguards that limit the spread of problematic updates.
The outage highlighted the extent to which many online services depend on Cloudflare for security and traffic delivery. Although the problem was resolved within hours, the incident showed how a small internal error can create a large impact when it reaches global systems.