It seems it's not the matter of if a cloud will go down, but when. Google and Amazon already have had well-publicized outages of their cloud platforms. Now, Microsoft has joined them, as Windows Azure was down for about 22 hours beginning Friday night.The outage, which according to a forum post by Microsoft employee and Azure team member Steve Marx, affected "a large number of deployments." Even after Microsoft identified the problem, the company took hours to restart the servers and get them online and running properly.

In his post, Marx said Microsoft would likely release an analysis of what went wrong, but not until after the company's MIX Web developer conference this week. Microsoft is widely expected to use MIX as a platform to make a few Azure-related announcements, though this is one topic the company would likely rather avoid.

Azure isn't yet a final product, and therefore doesn't have guaranteed service levels. Microsoft has said that Azure will come with strong service level agreements when it is finally released, and will likely use the outage as an opportunity to assess system recovery processes and figure out how to strengthen Azure's stability.

Though few developers and companies are yet making heavy use of Azure as a production system -- it isn't due for commercial release until the end of the year so it doesn't yet come with official support -- Azure's downtime serves as a gentle reminder that companies and individuals should make sure they have back-up plans in place in case something goes wrong with their cloud deployments. It's probably worth consulting our guide to cloud computing one more time.