Even at Cloudflare scale, an invalid useEffect dependency can bring everything down.
The gist of this article is that a bad dependency array of a useEffect caused excessive API calls which led to an outage. Reading about their strategies for mitigating this I was wondering why there is no mention regarding catching the original issue. It’s all about scaling better and rolling back. But why not also handle such invalid useEffect arrays. One idea I recently stumbled upon was this eslint plugin (react-you-might-not-need-an-effect). Great name btw!
Reference: A deep dive into Cloudflare’s September 12, 2025 dashboard and API outage