Five ways to get cloud computing wrong

Don't let your migration projects go up in smoke. Steve Cassidy runs through the blunders to avoid

- Steve Cassidy

Cloud computing is everywhere. Everyone’s doing it, and you can bet that a whole lot of people are getting it wrong in some way.

Before we look at how, though, let’s start off by making sure we get the definition of “cloud computing” right. It matters, because you can’t hope to have a successful relationship with the cloud if you don’t know what it actually consists of.

To be fair, the general conception of the cloud has been fuzzy ever since services such as Amazon’s EC2 started marketing to non-technical users. Prior to this I had been present in technical meetings and conferences where we all understood that “the cloud” meant a particular set of architectures presenting a particular set of services. But once marketeers started trying to sell the idea to managers, a more nebulous vision of the cloud began to take hold, as an all-capable, omnipresent yet invisible fabric that somehow runs the internet and anything else you want it to.

In fact, you can mostly understand cloud computing as a combination of three technological concepts; namely, hosting, virtualisation and DevOps. There’s nothing magical about any of them, but when combined in the right ways they add up to something a lot more powerful than their individual parts. It’s called “the cloud” not because it’s insubstantial, but because to deliver a scalable, managed, flexible hosting service, the technology draws on a vast number of fungible, commoditised servers, no different from one another than raindrops in a... well, you get the idea.

With a proper sense of what the cloud is, we can start to understand how some businesses get it wrong. The mistakes detailed below are all ones that I’ve observed in the wild – although they’ve been diplomatically anonymised to protect the guilty.

Mistake 1

Unrealistic targets