Yesterday, about 10.2 MILLION blogs fell off the face of the planet for just Under 2 hours when WordPress.com experienced it’s first major outage in four years. In all, about 5.5 million pageviews lost in those 110 minutes of downtime.
According to Matt Mullenweg, the downtime was caused by an unscheduled “core router” change at one of WordPress.com’s datacenter providers. I find it fascinating that a core router change at one data center broke failover mechanisms that should have shifted services over to another working DC. True to Automattic’s slogan, “Inspired by you, Striving to create good, Nothing is perfect.”, it seem someone was less than perfect in designing their failover system… and then forgot to test it.