The IaaS Revolution: Report from the Frontlines

When we read about revolutions historically, they are invariably painted as a momentary epoch during which the world blissfully changes forever and where the victors relish in the glory of positive new direction. Granted, revolutionary ideas are often harmlessly hatched in pubs, coffee shops, online groups, and simple meet-ups. But the revolution itself is anything but blissful or momentary. Revolution is about upheaval. And upheaval only brings uncertainty, panic and divisiveness. Revolution is a mess. It’s ugly and unorganized. And the business of it is bloody. If you want to storm the Bastille you better be prepared for bloodshed because the incumbent regime is not going to go down without a fight. And what’s worse, chances are there will be a cadre of competing revolutionists all bent on their own version of the ideal (Lenin, Stalin, anyone?). You need to watch your back as much as you keep your eye on the prize. Trust me on this one.

The technology business is no different. The grass roots movement to shift the way in which IT is delivered – from single tenancy to a multi-tenant utility – has created an intense battle. The specter of IaaS looms large over the IT industry and nobody in the supply chain is truly immune from its impact. You can see it everywhere you look in the business. Lines are being drawn. Alliances are being formed. The existing regime is deflecting and dissecting. And in the quest for customers (you know, those people that ultimately decide our fate) it is getting ugly. Revolutionists and incumbents alike wantonly steal ideas, poach people, shamelessly rebrand (pretty sure even my car dealer sells cloud now), undercut pricing and might even engage in a little espionage as they ruthlessly jockey for position.

So why revolt? Why do we put ourselves through such madness?

It’s simple really. We initiate change because the will of the people (or the market) demands it. Be you a disenfranchised member of the Proletariat that’s had just about enough of your bourgeoisie counterparts or a private sector business that’s had just about enough of hardware and software lifecycle cost the impetus is the same. We simply get to point where the status quo must go (ok, get your sign, we’re about to picket now!).

So how do we know when the technology revolution is over? Essentially, it will conclude when the market rate of adoption topples the existing regime and the floodgates open. Moore calls this “crossing the chasm.” I call this “the reason I get up in the morning.” It’s basically that point on the technology adoption curve that starts the hockey-stick (eh) looking upward slope.

Despite what you heard at the latest cloud conferences (these things look more and more like political rallies lately, don’t they?!?), IaaS is not yet ready for prime time.

Allow me to elaborate.

I see three common characteristics among every technological revolution that eventually gave way to a brand new paradigm in consumer society. It helps to think of these characteristics within the context of something everyone knows. You can pick pretty much any technology, but let’s consider three very different omnipresent exhibits to illustrate my point: The internet, electricity and the automobile.

Characteristic # 1: Global reach. I can transmit email around the world with the click of a button. I can consume electricity anywhere in the modern world. A car is an acceptable mode of transport pretty much anywhere. In order for a technology to achieve the upward slope of adoption it must be able to scale the globe in a uniform fashion.

Characteristic # 2: Technological abstraction. My 80 year old mother can send an email and she has never even heard of IP. I am not an electrical engineer, yet I can use electricity. My car is a complex piece of machinery, but all I really need to understand are a handful of simple instruments to competently operate it. It is impossible for any technology to permeate society if it is not made simple enough for anyone to use.

Characteristic #3: Universal measurement. A kilowatt is a kilowatt no matter where we live. Bandwidth dictates how many IP packets I can transmit as defined by every service bureau. Gasoline universally powers my car and I pay by the gallon/litre. Imagine how difficult it would be if we had to calculate octane levels and put out an RFP to Shell and Texaco before we made a decision to fuel our cars? Where would the internet be if every ISP had it’s own version of a packet? Where would we be if you had to manually calculate volts and amps for every device in your house to guesstimate your electricity consumption?

Now translate this to our very own IaaS revolution:

• IaaS today is regional at best, not global. Can I build a workload and deploy it anywhere I want in the world, from anywhere I want in the world? No. IaaS clouds are being stood up in silos by software peddlers capitalizing on hype and FUD.
• Can the average business user tap into any of the pioneering IaaS platforms and completely self-serve their computing needs? No. Using IaaS is complex and limited only to engineers and sophisticated software developers.
• Can I measure my compute utilization the same way whether I operate it in Chicago, London or Tokyo regardless of the local service bureau? That’s a pipe dream for most customers and even service providers. IaaS metering and billing is painfully done in private sandboxes and vacuums today.

The next phase of the revolution is upon us. As reach, abstraction and measurement become more uniform; more pervasive, simple to use, you will begin to see the light at the end of the path.

Take it from a self-proclaimed revolutionist: Only then will a truly new era of IT delivery emerge.