The Government Forecast is Mostly Cloudy

In what may be the most significant move in the history of IT, America is moving to the cloud.

Follow the money and you'll see what I mean.

The U.S. government spends billions and billions of dollars on computing each year. When new CIO Vivek Kundra took office he found $27 billion in projects that were either behind schedule or over budget.

Now, just as the media is hyperventilating over Amazon's FAIL with EC2 over the weekend, Kundra is being profiled as the man behind a cloud first strategy that will close 800 data centers.

So what's going on?

Amazon is not the only cloud out there. It is a vendor of Infrastructure as a Service (IaaS) services. It is not “the cloud.”
Clouds are really just a more efficient way to run IT infrastructure. Any enterprise that doesn't embrace efficient is a dumb enterprise.
Still, there are a ton of large federal contractors who are threatened by the move toward a cloud paradigm, and all of them are quietly spreading FUD about reliability, security, and privacy in order to slow this train as much as possible.

It may in fact be coincidental that the Amazon cloud problems began in a data center located in northern Virginia, where many of these contractors are based. Amazon said the problems involved latency and connectivity, which sound more like networking issues than software. As of now things are back to normal.

Still, it's clear that cloud computing is going to have to climb a wall of worry, or at least a wall of stupid. As with Google last week, Amazon is expected to suffer at the hands of investors for daring to invest in its infrastructure.

It's crazy. As the federal government moves forward with its cloud strategy, there are going to be hundreds of billions of dollars in play, much of which will go to companies who already have something to sell. (Some, admittedly, will be wasted by contractors using Uncle Sam to get into the game.) Investing in clouds is the biggest no-brainer in the history of tech.

Clouds are a technology. They're not a place, not a system. They're a way of doing things. Open source is going to have to get a big play in it because clouds are complicated and designed to be shared, which you can't do unless the software is visible. Even Microsoft has gotten religion on this with its Azure cloud.

The biggest mistake we can make in clouds today is to over-think them. They're evolving rapidly just as any computer technology evolves, with a lot of players jostling for minor advantages. But this is the first enterprise technology to evolve completely within the open source era, and built by companies with both open source heritage and understanding.