Cloudflare says it has fully resolved the global outage that triggered widespread 500 errors across the internet. Platforms like ChatGPT, X, Perplexity, Canva, Uber, League of Legends, and Valorant all went down, alongside Cloudflare’s own dashboard and API. CTO Dane Knecht called the incident a “failure” and “totally unacceptable,” confirming the cause was a latent bug triggered by a routine configuration change not an attack.
While Cloudflare outages usually last seconds or minutes, today’s disruption stretched far longer. Even so, considering the scale of their network and how many daily changes roll out silently without issues, the recovery time was still impressive.
And this isn’t just a Cloudflare story. AWS suffered a major outage last month, taking down Snapchat, Reddit, and several large platforms. Amazon apologized and acknowledged the increasing difficulty of maintaining flawless uptime in sprawling, interconnected cloud systems.
Both incidents underline a hard truth: modern infrastructure is powerful, but fragile. No provider not even Cloudflare or AWS can promise 100% uptime. What matters is how quickly they isolate the failure and restore the world’s digital pulse. On that front, Cloudflare delivered.
