When is a cloud not really a cloud?

...When you stick the word "private" in front of it.

I have a pretty cool data center. State of the art, IP storage, Virtualization, you name it, we are fully buzz-word compliant. When I was describing to some of my peers at a CIO conference not long ago, one of them said "You have a private cloud!" with much enthusiasm. It took me by surprise, as I had never thought of it that way before. I said that I guess you could call it that. Later, after much discussion and wrestling with the idea of a private cloud, I decided there was no such thing. Sure the modern virtualized data center has advantages over the old physical world, but it also comes with its own set of challenges, and my internal IT group still has to deal with physical infrastructure, and more critically, maintain and upgrade apps.

With my cloud applications (the real ones) I never have to worry about the apps, they are just there. Most of them give me cool new features, stuff that adds value to the business, multiple times per year. In contrast, we just spent many man-months upgrading our MS Exchange system. Sure there are a few cool new features, but if I had put that much energy into adding new business value, it would have been much better spent.

I still maintain that I am not moving to the cloud for clouds sake. I am solving problems for the business. Most of the new things we do these days are in the cloud, but we choose them because that seems to be where all of the innovation is these days. When was the last time you really saw something innovative from one of the traditional big software vendors?

And on a similar note, taking you bloated, traditionally in-house application and running it on public infrastructure is not the same thing. Using my Exchange example, you'll still wait three years between batches of new features.