Cloudflare: Last night's widespread outage was due to an internal configuration error, accelerating our system's resilience upgrade

BlockBeats News, November 19th: Cloudflare released an analysis report on the large-scale outage on November 18th, stating that the most severe outage since 2019 was caused by a misconfiguration generated after a database permission change, leading to an abnormality in the core proxy system, affecting multiple products such as CDN, security services, Workers KV, Turnstile, and Access.

The team initially misidentified it as a DDoS attack and ultimately resolved it by rolling back to an old file. All services were restored at 01:06 on November 19th, Beijing time. Cloudflare stated that such outages are unacceptable and they will accelerate system resilience upgrades.