Overview
- The company reported that a configuration change at 05:05 local time caused duplicate database rows and generated a feature file that exceeded the Bot Management module’s limit, crashing request handling at its proxy layer.
- Automated tests flagged anomalies by 05:31, an incident call was opened at 05:35, major impact was reduced by about 08:30, and full restoration was declared around 11:06.
- User-facing symptoms included blank pages, internal server errors, and Cloudflare challenge prompts, with outage spikes recorded by Downdetector and other monitors.
- X, ChatGPT, Spotify, Canva, League of Legends, Shopify, Dropbox, Coinbase, Moody’s, and transit and media sites reported disruptions as Cloudflare rolled back changes, halted new Bot Management file propagation, and restored products like Access and WARP.
- Executives apologized, called the outage unacceptable, committed to a technical post‑mortem, and external analysts underscored concentration risks given Cloudflare’s handling of a significant share of global web traffic.