I use Microsoft Azure Traffic Manager to fallback to a 2nd server if my main server is unreachable. I use Pingdom to monitor the main server availability. Usually it's 99.xx depending on Windows updates and how much work I did on the server. This is not a problem since the 2nd server will cover this. But in the last two couples of months, I've got my worst availability followed by the best one!
The downtime is caused by Azure shutting my services because I've reached my spending limit.
