It used to be the case that lines of business had no choice but to wait in line to deploy applications into production. Waiting weeks, months or longer caught in this “deployment gap” has never been a particularly welcomed fate, but it has become the accepted reality for lack of better alternatives.

But this is beginning to change, as public cloud services emerge as an “escape valve” for the classic IT bottleneck. Rather than waiting in vain for their number to be called, business lines have what appears to be a potential alternative. Since enterprise applications are typically driven by specific and often very strategic business needs, this lower-friction alternative is particularly welcomed.

Of course, I’m oversimplifying the reality that some applications can’t —or shouldn’t — be deployed into a public cloud environment. But that doesn’t mean they won’t. I believe many applications will follow the path of least resistance, escaping to the cloud based on the swipe of a credit card and a philosophy that begging forgiveness can be more efficient than asking permission.

I’ll bet dollars to donuts these rogue applications are escaping to the cloud today.

The reality is that delay often means missed opportunity. This is particularly true in the context of applications that are driven by transient demand; that is to say, the need is here today, gone tomorrow, and it can’t be readily predicted. You see a lot of this in financial services and other domains where applications are deployed to execute massive-scale computations in response to some dynamic need — for example, econometric models to exploit an arbitrage opportunity. Another example might be natural resource discovery, where the same sort of large-scale simulations may be run based on what appears to be a potential oil deposit.

When the need is ephemeral and the stakes are high, waiting isn’t a valid option. For business lines, cloud is becoming that shiny object that says: You can have what you need, when you need it. It’s an enticing proposition for those with a need for speed.

Of course, this is all putting backpressure on IT operations to find a way to say “yes,” when the default answer has typically been “no.” Cloud is creating a (sometimes unwelcomed) cultural change for IT operations that have seen speed as the enemy of control. Cloud — particularly public cloud — has made this conflict very clear.

Not unlike toddlers and cats, the business side can be a willful bunch. Saying “no” will not tamp down rogue applications escaping to the cloud. The better strategy is for IT to find a way to say “yes” by providing a managed, policy-driven approach for packaging, deploying and managing applications in the cloud.

This managed approach delivers both business lines and IT what they need, blending the virtues of speed and control. For IT, it provides a far more efficient way to align capacity to demand by allowing infrastructure to become truly elastic. In the case of external clouds, it may even provide a way to deliver compute horsepower to massive-scale, but transient, application workloads without ever building out another datacenter.

So, the moral of the story is that rogue applications will escape to the cloud — and they probably already have. The solution isn’t to prevent it, but to control it.

Organizations that institute policies for governing how, when and where applications get deployed to the cloud will find themselves in far better control than those who attempt to snuff it out by way of restrictive edict. That’s the paradox of this story: Saying “no” to cloud may well become the greatest risk of all.

You do not have flash or javascript support.

Average (1 vote)

Reply