Are you ready for the coming zombie apocalypse?

Not flesh-eating mutants who wander the earth in search of fresh brains, but something far more horrifying and insidious: the living dead server.

Look around, and you’ll see it’s already happening. Companies are spending millions and millions on virtual machines (VMs) and many show no real signs of visible life.

The saddest part is that most organizations aren’t even aware they’re throwing away valuable resources and dollars.

Redundant People and Machines

Perhaps even more frightening: in this new, post-infrastructure era, there will be a lot of IT casualties beyond under-utilized machines — the new walking dead.

We’re entering a resource stratum in which people and machines are becoming redundant.

Businesses are beginning to replace people trained to operate in a universe where IT is an automated commodity with fewer people who can function more nimbly in server-less architectures.

But this shift puts significant pressure on day-in, day-out management. When you move from a capital expenditure (CapEx or upfront) model to an operating expenditure (OpEx) approach, your needs change dramatically.

Things like auto-scaling, scheduled operations and utilization become hugely important and provide the foundation for a new class of IT operations and staffing. Quite simply, you need an entirely new operating model for the cloud.

The Dawn of Smart Managed Services Providers

Amid this two-pronged challenge, companies are learning valuable lessons.

Rather than turning to different managed service providers or operating models, they’re finding one, smart managed services provider can do the trick — from migration to operations, with a uniform set of standards, best practices, staffing and tooling.

From a responsibilities perspective, the onus is on the consumer of cloud services (not the underlying hyperscale player) to provide the required operational integrity.

Frankly, Amazon, Google or Microsoft don’t care how much you spend each month — but obviously you care.

Consider IT Complexity

The secret is consistency.

If you start down the path with a disjointed approach, write some scripts and cobble together an ad-hoc solution to solve for a very specific aspect of a specific problem, you may solve for the immediate challenge, but you will not realize that you had another 100 problems queuing up.

Similarly, adding a second and third provider exponentially increases the magnitude of your IT complexity.

You may have hardwired your automation to a provider template, but later realize the work isn’t extensible to other providers.

The danger in both of these situations is that many times, problems don’t manifest until you have a few hundred VMs under management across different domains or business units. Problems then multiply, and costs skyrocket.

Look at the Big Picture

To combat this stitched-together, fragmented approach, companies should consider their IT needs holistically — not just at one problem at a time.

Think big picture. How is your business going to advance from point A (legacy IT) to point B (the cloud) and manage more nimbly and efficiently once it gets there (points C through Z)?

What does your world look like when you’ve migrated 50 percent or more of your estate?

In other words, smart companies don’t Band-Aid each little IT issue they find.

Instead, they take a much more strategic approach to IT. That’s a tall order, of course, when you spin up a few hundred VMs simultaneously.

Again, that’s how problems you didn’t even know you had can creep up from behind. It’s also why I caution against a boutique or one-off, point solution.

Unwieldy Beasts

Most cloud providers’ bills are simply unmanageable — they’re millions and millions of lines of often unintelligible, raw data. You could be spending $30,000 a month on cloud or $100,000 —and really have no idea.

Meanwhile, your CFO is breathing down your neck because you’ve blown an entire year’s budget on two months’ worth of IT.

And you didn’t realize you had a problem until well after the month ended. You’ve moved from a controlled CapEx environment to an uncontrolled OpEx environment in which time and visibility are much more critical.

The world we’re operating in is no longer one in which you buy cement and boxes, and depreciate and amortize those assets. This is a very different space. Typically, old IT was purchased for peaks, whereas cloud is purchased for valleys and peaks, spun up and down as needed.

With the proper managed service provider, you’re buying a package of best practices, policies and permissions to run your new IT estate. The provider should offer the skills, methods and tools to help prevent sticker shock (among other things).

We are living in the post-infrastructure era (akin to how Steve Jobs and Apple helped usher in the post-PC era). In this brave new world, companies that don’t combat the Zombies with the latest tools, technology and talent will surely succumb in the coming apocalypse.

Zombie Watch

Smart cloud platform managers can help you save costs and increase your IT estate-awareness with tools such as “Zombie watch.”

Because of the unwieldy nature of VMs, you can control costs more precisely by knowing which machines are essentially going to waste — and prevent sticker shock when you receive a bill.

If you have a zombie watch feature built into your cloud management tools, you can more precisely gauge CPU usage (and therefore costs).

It’s an intelligent, dynamic way for businesses to catch the most pernicious problems (particularly wasted CPU space) before they arise, and kill a zombie before it kills you (and your budget).

Title image "Zombie" (CC BY 2.0) by EvanManphis