Cloudflare pinpoints system glitch that knocked major websites offline worldwide
Technology
Cloudflare identified an internal bot-management system glitch — triggered by a database change — as the cause of a global outage that affected major websites like X
SILICON VALLEY (Dunya News) – Cloudflare, the global cybersecurity and web-infrastructure giant, has traced Tuesday’s massive internet outage to an internal glitch in its bot-management system – the very tool designed to keep traffic in check.
Matthew Prince, Cloudflare’s founder and CEO, said the disruption wasn’t caused by a DNS failure, a new AI tool or even a cyberattack. Instead, the problem boiled down to an internal database change that went wrong behind the scenes.
Cloudflare, which handles nearly 20% of the world’s internet traffic, said its bot-management system uses a machine-learning model that assigns a “bot score” to every incoming request to filter real users from automated crawlers. The model relies on a configuration file that is frequently updated.
But after a change in the ClickHouse database, the system began generating duplicate rows, causing the configuration file to grow far larger than its allowed limit. Once the file exceeded its memory cap, the core proxy system, responsible for managing web traffic, mbegan to fail.
As a result, traffic from legitimate users was blocked, knocking several major platforms offline for hours, including X, ChatGPT, DownDetector and others. However, systems that did not rely on the bot-score feature remained unaffected.
