Over the weekend, power problems caused the majority of Wikimedia projects to be unavailable for part of a day. The problem, traced to a technical failure at the Florida colocation center, hit on Sunday and took a little over six hours before service could be restored.

On 9 April at 19:19 UTC, the majority of Wikimedia sites, including the English Wikipedia, became unavailable due to the outage. Chief Technical Officer Brion Vibber soon reported that "PowerMedium had some sort of network and/or power outage for about an hour. We've been working on getting things back online since power and network became available again." He later clarified that "an 800-amp breaker in the main PDU failed".

Developers along with colocation center staff continued working to restore service, making sure that things could operate normally again. Service was restored at 1:33 UTC on 10 April, after six hours and 15 minutes of downtime. Vibber reported that some of the servers came up with bad clocks when power was restored, so some users might have seen watchlist entries reflecting incorrect times.

During the outage, the #wikipedia IRC channel was flooded with users seeking an explanation, peaking over 1000 people at one point. Rob Church reported that the developers had to set the #wikimedia-tech channel (devoted specifically to technical discussions about the servers) to moderated status in order to be able to work effectively and not constantly be interrupted by the same questions.

A similar power failure at the colocation center occurred in February of last year, with a smaller equipment failure following in March.