Albert Einstein said insanity is doing the same thing over and over and expecting different results. By that definition, many IT departments are insane.
Harsh, right? Well, the truth hurts. The IT culture is to talk ad nauseum about change, new methodologies, cloud, converged infrastructure and so on. People in IT will read The Phoenix Project, download a Puppet Enterprise trial and even buy a few hours of services in the public cloud. They will tell others that they're using agile methodologies to manage their IT projects, say they're amidst a configuration management rollout and brag that they're using the public cloud.
But they're not. And on the back end, where IT culture prevails, they continue to squander opportunities for change at the end of equipment lifecycles, reinvesting in the technologies and methodologies that created the problems in the first place. They're buying monolithic and expensive storage arrays and connecting them to their antiquated Fibre Channel storage area networks. They're building servers and infrastructure by hand, or using highly customized templates rather than moving forward with configuration management systems such as Puppet or Chef. And they're telling people that they're ready for the cloud, because, after all, the cloud is just another data center somewhere else. Right?
Many people would consider all of this to be lying, and I used to be one of them. But I don't think they're lying, anymore. I think they're insane. They do the same thing over and over, with the hope that maybe this time it'll be different. Maybe it'll be cheaper this time, or more resilient. Maybe it'll be more aligned with business needs, or somehow require fewer staff to maintain. Of course, the truth is closer to the opposite of all of these things. Unless someone in IT makes a conscious decision to do things differently, IT culture will not change, but the fears of risk and change keep everybody trapped in 1980s technology and processes.
So how does an IT group un-trap themselves, or get themselves unstuck from the predicament they're in? How do they change their IT culture? Because there is never a good time to start, they'll tell you. They're always busy. They're always in the middle of a giant project. They always have an excuse for why they can't do it. Unless they make time, though, it'll never happen, and without a goal to achieve there is never going to be forward movement. So set some goals and make some of these changes your top priority. And yes, this might mean telling other projects that they're not as important. Remember that a rotten foundation cannot support great works, no matter how important those great works are.
IT must also admit that very few of its staff are ready or willing to change the prevailing culture and mind-set. It's hard to realize that the longer someone is in IT, or any profession, the more set in their ways they become. Those ways might be expensive and time-consuming, but the folks that are stuck won't get fired if they just stick to the tried-and-true plan they came up with in 1997. They're probably right -- they won't be fired, they'll be laid off when the company is acquired for pennies on the dollar.
Change is messy. Surrender to it, embrace the chaos. What works in many cases is to create a tiger team or a task force: a small group of people given requirements, a timeframe, some budget, but also free reign, to create something new. These teams go out and recreate IT without the baggage of the rest of IT, and if it works they get reintegrated into IT as a whole. Or, more pointedly, IT gets reintegrated with them, under their culture, methods and tools, in a more formal, non-chaotic fashion.
You cannot finish what you never start, and to change you must actually make changes. Give it a shot. Buy a hyper-converged compute platform. Start deploying apps and services with Puppet. Move some things you care about to the public cloud. Only then will you be making meaningful changes, ones you can really brag about to your friends and colleagues. You'd be crazy not to.
Bob Plankers is a virtualization and cloud expert at a major Midwestern University.
Dig Deeper on Data center budget and culture
Research: Cultural barriers hold enterprises back from building ‘high-performing’ DevOps teams
Open source licence series – Puppet: consumption without collaboration equals consternation
Why enterprises should care about DevOps
Transform network management with a network automation strategy