All of our websites went down on Jun 21 04:36 UTC (9:26 PDT) due to a major outage in the Linode Fremont data center. I gave Linode an hour to get things back up, and when things were still down I started to rebuild the web cluster in their Dallas data center. Service is restored as of 6:30 UTC (11:30 PDT) for a total outage time of just over 2 hours. I'll be monitoring this and will shut down the extra cluster as soon as they have things working again.