There’s no future in on-premises IT — it’s time to move to the cloud

I’ve been lying to myself: I thought IT would survive the next shift in technology as all infrastructure moves to the cloud. But I no longer believe IT will survive that cloud shift — certainly not IT as we know it. Sure, there will always be on-premises dinosaurs like myself who prefer to install Exchange manually. But the shift to the cloud is coming.

The debate between operating expense (aka opex, the cloud’s approach) and capital expense (aka capex, the on-premises approach) is waged daily at companies. Although there are trade-offs no matter what a firm chooses, it’s clear that opex is increasingly favored. In the age-old rent-versus-buy debate, the cloud is making rental very compelling, especially as managed cloud environments begin to implement tools that provide automated spin-up/spin-down to avoid excessive consumption of resources (and higher costs).

I recently read “The Big Switch” by Nicholas Carr, the same guy who wrote the controversial 2004 title “Does IT Matter?” that questioned the future relevance of IT. His personal opinion aside, what struck me about “The Big Switch” is Carr’s explanation of how the use of electricity shifted from on-premise systems — companies had electricity departments, complete with electrical architects and managers, sort of like IT departments today — to third-party electrical grids that businesses simply tapped (like today’s cloud services). Electricity went from an item on which a business focused half its time, attention, and labor to a simple utility it plugged into and paid for. Like it or not, the same is happening in IT.

Some of us remember the days when IT held great prominence in any business and, as a result, saw a lot of money sent its way. But 20 years later, businesses sign up for Office 365, and an office manager adds user accounts to a simple Web interface. What was once handled by IT is becoming a commodity service provided by something very much like a public utility.

I recently logged into a Windows Azure portal, picked a template that included Windows Server 2012 and SharePoint, and spun it up in minutes while I went and got a cup of coffee. It was finished well before I returned.

I could have done the same thing on-premises too, but not without an infrastructure and a good deal of time to get the software in place to make it happen. By contrast, with Azure I didn’t have to do much of anything but choose what I wanted. It doesn’t matter whether you use Azure or another vendor’s service (I opt for Azure because it works with the Microsoft technologies I already know); the patterrn is the same: Log in, make a few choices, and you’re done. That’s not the future — that’s the present.

If you started in this profession in the late 1990s as I did, you might tell yourself, “I’m going to be that on-site expert that will always be in need.” You might actually make it through to retirement with a few PowerShell classes, some automation experience, and a few sessions at a conference about System Center.

But if you’re looking for any longevity in your IT career, you need to move past that kind of minimal coping. The shift to a full cloud-based infrastructure-as-a-service (IaaS) or platform-as-a-service (PaaS) model is imminent. Yes, in the next five years there’ll be hybrid and convergence solutions teed up across the board to provide a transition from on-premises to cloud, but ultimately there’ll be little on-premise IT left for admins.

I recently asked a room of IT pros how many of them had worked with on-premises virtualization for their servers. They all raised their hands proudly, as if they were “modern” and not living in the last decade. Then I asked how many have used a cloud-based service like Azure for testing or production servers. Only two out of 30 people raised their hands. That shows the risk most IT admins face today, whether they know it or not. For IT, 2000 to 2010 was about virtualization, but that’s done. It’s no longer modern, simply the new legacy. The current decade may be focused on convergence and hybrid models. After that, it’s all cloud from what I can see.

To avoid irrelevance, you must change and grow. You can dig your heels in and cling to the past, as I have done often enough, but that will not help you once cloud-based tools overcome the increasingly addressed concerns over security, availability, performance, and flexibility.

You won’t be the first to make such an adjustment. Thomas Edison, who invented the earliest system for electricity distribution in 1880, initially used direct current (DC). He didn’t want to adopt alternating current (AC), though it could use higher voltages with transformers to step it down for distribution to homes and businesses. Competitors like Westinghouse were all about AC, but Edison doubled down and waged a propaganda war against the technology, trying to ban AC — even electrocuting animals publicly to demonstrate the danger of AC. Ultimately, as we know, he failed, and AC current became our norm.

Once you see technology going in a direction that’s unstoppable, don’t get in the way or try to stem the tide. Instead, get on board. The cloud train is here.

This story, “There’s no future in on-premises IT — it’s time to move to the cloud,” was originally published at InfoWorld.com. Read more of J. Peter Bruzzese’s Enterprise Windows blog and follow the latest developments in Windows at InfoWorld.com. For the latest business technology news, follow InfoWorld.com on Twitter.