Like past AWS outages, two physical locations (Dallas and Chicago) going down at once. Even if it were feasible to set up async replication to the London location, those servers are on a different platform altogether (not workable through the US Cloud Servers control panel and API). So I guess Rackspace isn't really an option for HA. Not that AWS looks much better ATM.
My server just came back up - 4h30 down. A lot of fail in this one, with status pages only reflecting issues 50 minutes later than my monitoring system, and phone techs claiming "no official information but huge call volume" after being on hold for 20 minutes.
