I recently wrote a new Yankee Group Decision Note titled “Considering Cloud Computing for the Anywhere Enterprise” (it isn’t on the Web site yet, but will be soon). In that note, I warn early adopters of cloud services — even those from such tried-and-true providers like Amazon.com — that they should understand five different risks, one of them being reliability risk, before they jump in with both feet.

That note feels a bit prescient today because Amazon.com’s main site, www.amazon.com, has been struggling to provide Web service today, instead providing error messages like the one shown above (click on the image for a larger version). And while I don’t think this has propagated to the Amazon Elastic Computing Cloud service, it does argue that even the top tier providers are a long way from a perfect record in providing non-stop service.

The bottom line:  transferring an IT function  — whether it be hosting a Web site, running a CRM system, or managing a data center — to someone else doesn’t mean you don’t have to worry about it. It just means you won’t be able to fix it yourself when it goes wrong.