A cooling system failure at a Phoenix data center knocked out services for two major hosting providers at once, and the outage says something more interesting than “servers went down.” It reveals how much these companies actually depend on a single physical building.
The problem started August 13 at a facility where phoenixNAP operates as a tenant. According to status updates, elevated temperatures followed storm-related utility disturbances overnight. RadiusDC, which runs the facility, brought one chiller back online and rolled in temporary cooling equipment while working to restore the rest of the system.
Namecheap felt it across nearly everything: shared, VPS, and dedicated hosting, EasyWP, private email, DNS management, forwarding services, even parts of its own website and support helpdesk. The company had to post backup Microsoft Teams contacts just so customers could reach someone. Hosting.com reported servers unavailable at its own Phoenix location, confirming this wasn’t a Namecheap-specific problem. It was the building.
That’s really the uncomfortable part for anyone thinking through infrastructure risk. Two separate companies, different customers, different systems, still went down together because they shared the same cooling plant. Commercial separation between hosting providers doesn’t mean physical independence, and this outage made that distinction impossible to ignore.
Namecheap noted that DNS resolution stayed up even while DNS management didn’t, a detail that actually matters. Losing the ability to manage your DNS settings is different from losing DNS entirely, but neither is exactly reassuring if your business runs through it. The company said it was restoring physical networking first, then virtual infrastructure, then customer-facing systems, a staged approach meant to avoid a second failure from bringing everything back online at once and overloading power or storage systems.
There’s also a business context worth noting. RadiusDC only agreed to acquire this facility from phoenixNAP back in March, with phoenixNAP staying on as a tenant. An outage this early in that transition will inevitably draw scrutiny toward the site’s cooling redundancy and how it holds up under stress from utility disruptions.
For any business running hosting through a Phoenix-based provider, the outage raises a fair question: whether internal redundancy at a single facility was ever a real substitute for actual geographic backup.
