Lies, damned lies, and cloud statistics
I'm a realist. I don't pretend that things are perfect when they're not. I don't ignore problems. I don't try to tell people that OpenStack, or even private cloud in general, is perfect.
But let's get real, people.
Last week, everybody's favorite OpenStack character assassin, Gartner, came out with a survey that "proved" that 95% of private clouds are failing. In this survey, they asked the question "What is going wrong with your private cloud?" and were "a little surprised that 95% of the 140 respondents (who had private clouds in place) said something was wrong with their private cloud."
"Something's wrong" does not mean "failing"
Right there at the top we have a problem. Just because something is going wrong doesn't mean that a private cloud is failing. Let's look at their results:
Take, for example, "Failure to change the funding model"; a cloud can be wildly successful at providing benefit to the business, but nobody's figured out who's paying for it. Or "Focusing on the wrong benefits"; maybe your developers are suddenly 500% more productive (since they don't have to wait for resources) but you're focused on the fact that it's still costing you money to run the cloud. Maybe your problem is a "Failure to change the operational model" and you have a could that's doing fine -- but could be spectacular if you could just get your people to move to an agile model, or use DevOps.
My point here is that just because something can be better doesn't mean that it's "a failure", and it's irresponsible -- or at the very least, disingenuous -- to say so.
What's going right?
The fact is that I could just as easily come up with a chart that says that 95% of private clouds are successful. Since I'm making a point about semantics here I don't have time to survey 140 people but let's assume that I asked them, "What benefits are you seeing from your private cloud?" I might come up with a chart that looks something like this:
If I used the same "methodology" as Gartner, this chart would "prove" that 95% of all private clouds were successful.
What Gartner's chart is really telling us
All of this is not to say that the Gartner survey isn't useful; it is. It tells us where private cloud -- and more specifically, companies using private cloud -- can see improvement. Some takeaways from this survey are that companies should think about include:
Doubling down: If your cloud isn't doing enough, get out of "proof of concept" mode and get into production.
Reviewing financing: Should you be funding your cloud in more innovative ways, such as OpEx vs CapEx, or being more specific about chargebacks and monitoring?
Stop defending Infrastructure and Operations: Maybe you should consider Hosted Private Cloud or Hybrid Cloud rather than that massive private cloud you're struggling with.
Focus on automation and DevOps benefits: Cloud enables you to escape the tyranny of manual labor (and configuration) -- but only if you let it.
Change your operating mode: One of the main advantages of cloud is agility. Take advantage of it!
Private cloud is still a fairly nascent industry; it's natural for companies to take time finding their way, and no matter how hard traditional vendors try to fight it, it's coming. Make the most of it.