Overview
- The Nov. 18 incident disrupted access to many websites worldwide, with intermittent problems reported for some users of ChatGPT and X.
- Cloudflare reports that a change in database access permissions generated a faulty Bot Management configuration whose file size doubled beyond limits, triggering failures in traffic-handling software and client services.
- Engineers initially suspected a large DDoS attack before determining the issue originated in the internal configuration update logic.
- The company halted propagation of the bad file, rolled back to a stable version, and reported full infrastructure recovery at 17:06.
- Customers experienced elevated error rates during restoration and Cloudflare’s support portal and status page also faltered, underscoring the systemic impact of outages at a widely used provider with notable incidents in 2019 and 2022.