↓ Skip to main content

“Here’s a post we never thought we’d need to write: less than five months after one of our major data centers lost power, it happened again to the exact same data center.”

Matthew Prince, John Graham-Cumming, and Jeremy Hartmanblog.cloudflare.com ↗

Flexential’s PDX01 went dark again on March 26, when four switchboards failed at once and took out both primary and redundant power. The cause points to breaker settings set too low for the load. Cloudflare’s APIs came back in seven minutes this time, against six hours of control plane downtime in November. It still runs its control plane out of the same building, and analytics still went down with it.