by Bernard Golden

Cloud Computing Will Cause Three IT Revolutions

Feb 09, 2010
Cloud ComputingVirtualization

During the next two to five years, you'll see enormous conflict about the technical pros and cons of cloud computing. Three of cloud's key characteristics will create three IT revolutions, each with supporters and detractors.'s Bernard Golden explains the first of these, involving scalability and IT operations.

Every revolution results in winners and losers — after the dust settles. During the revolution, chaos occurs as people attempt to discern if this is the real thing or just a minor rebellion. All parties put forward their positions, attempting to convince onlookers that theirs is the path forward. Meanwhile, established practices and institutions are disrupted and even overturned — perhaps temporarily or maybe permanently. Eventually, the results shake out and it becomes clear which viewpoint prevails and becomes the new established practice — and in its turn becomes the incumbent, ripe for disruption.

[For timely cloud computing news and expert analysis, see’s Cloud Computing Drilldown section. ]

This is true in technology as in every other domain. In the tech business, we often get caught up in focusing on vendor winners and losers. Point to the client/server revolution, and it’s obvious — Microsoft and Intel. Over on the loser side stand the minicomputer vendors. This winner/loser phenomenon can be seen in every significant technology shift (and indeed, one shift’s winner can become a future loser). This is understandable: we all love conflict and the vendor wars make for great press.

Less awareness is present for the effects of these revolutions on what makes up the vast majority of the technology industry — users. One could hazard a guess that for every dollar of revenue that Microsoft products pull in, IT organizations spend 10 or 20 additional dollars (or perhaps even more) in building and running systems. By far the biggest impact of any technology revolution is that upon technology users (by which I mean those who work with the technology, i.e., IT organizations).

Another aspect of change is how individuals react to it. It’s a cliche that “people don’t like change.” That’s dead wrong. People accept — and even embrace — change when they see it brings a direct benefit. Look at the immediate adoption of the iPhone — didn’t see a lot of resistance to that, did you? A more nuanced understanding of people’s reaction to change would interpret likely reactions based upon how the effect of the change is perceived by the individual — is it a benefit or a threat?

When it comes to organizations, it’s a misreading to assume that the organization will react as a whole — every organization is made up of groups and individual actors, each of which will have its (or his or her) own read on the implications of a change. If we look to the original move of PCs into companies, some portions of IT organizations embraced them, while others, wedded to the existing mode of performing IT, saw them as a distraction, a threat, or a toy. In other words, there were different camps that arose in reaction to the availability of this new form of computing, and there were pitched battles for personal and organizational influence that took the guise of a technical assessment.

When it comes to cloud computing, we should expect to see the same dynamic play out. Over the next two to five years, expect to see enormous conflict about the technical pros and cons of cloud computing that will, at bottom, be motivated by the perception on the part of the participants as to whether cloud computing represents a benefit to be embraced or a threat to be resisted.

In particular, cloud computing’s three characteristics — the illusion of infinite scalability, lack of a long-term commitment, and pay-by-the-use — will result in three revolutions in the way IT is performed, and each of the revolutions will have its adherents and detractors.

Revolution #1: The Change in IT Operations

Much is made of the magic of Amazon Web Services — fill out a web page, hit a button, and 10 minutes later, you’ve got computing resources available. Even more impressive, you can obtain large amounts with that request. And later, should you need even more resources to be added to your original pool, they’re easily requested and joined to the existing resources. This is the vision that many find so tantalizing, given today’s lengthy provisioning cycle, which in many companies results in months-long gaps between request and resource availability. Many think removing all the friction of resource provisioning is a huge win. One might think of this change as the logical extension of the view that hardware has been transformed from a scarce, expensive resource into a cheap, easily purchased commodity — the logical outcome of which is the need to treat provisioning it like a mass good, not a precious luxury.

Perhaps less obvious is the implications of this vision — that existing processes and organizational structures need to change to support this new mode of automated management.

Today, IT organizations interpose a large set of processes and requirements in the provisioning process. Budget requests, discussions with the various operations groups like network and storage, scheduling meetings, all surrounded with lots of paperwork. And these mechanisms make sense for an environment in which they help ration scare resources. They are in place to ensure that each precious resource is devoted to its highest possible use.

The problem is that these mechanisms are orthogonal to the streamlined, short-duration provisioning associated with cloud computing (the au courant term is orchestration, representing the unified bringing together of resource assignment in an automated manner). In effect, there is an impedance mismatch between the operational implications of cloud computing and the organizational artifacts that exist today. And, as noted at the start of this piece, any time this kind of mismatch occurs, there is bound to be organizational conflict — carried on at the level of technical discussion. After all, no one is going to say about cloud computing, “I don’t like it because I’m not sure how my job managing the installation and configuration of servers will be needed when someone can just fill out a web form and have the infrastructure itself arrange for the provisioning.”

So how will this play out and who will be the winners and losers?

Winners: Apps groups. Apps groups are driven by business groups, many of which are frustrated by not being able to react to urgent business pressure. This is not to mention the frustration many feel when confronted by the “owners” of the resources who assert their judgment as to whether the request is justified. Bypassing all of this organizational overhead and being able to react much more quickly to business developments is a huge win. Expect to see enormous pressure from apps groups to “get on the cloud.” And if the operations groups don’t respond quickly enough, expect to see the apps groups look to outside providers which have a financial incentive to respond immediately. (I addressed this in my last post, here).

Winners: Apps groups (2). The high-friction provisioning process hasn’t merely been the result of rationing by the operations groups. This rationing process ends up being backed into the apps groups themselves, where different business applications vie to be put onto the request list. This has the inevitable outcome that many business applications never “make the cut” to be submitted for resources. And often, these are the applications that represent innovative but unproven applications of information technology. The process goes something like “well, we know we *have* to schedule the upgrade of the XYZ package, and we know we need to refresh the hardware that the ABC application runs on, so that pretty much covers what we can do this quarter. Bob, sorry that we can’t address your application that matches our customer complaints against our manufacturer partner’s trouble tickets to see if we can identify breakdown patterns.” Low-priority applications will have much more opportunity in a cloud computing world. A complement to this is the inevitable overall growth in the use of IT resources.

Losers: IT operations. Putting provisioning in the hands of IT resource users inexorably results in less influence for IT resource “owners.” As I was outlining this vision of end user-driven, automated provisioning in a workshop last week, one attendee said “I can see why it’s so attractive, but I can’t see IT operations accepting it. I think we’ll hear something from operations groups like ‘automated provisioning is great, but it should be done in my group to make sure requests align with accepted standards or the like.'” Of course. This is something like what ecommerce sites did 10 or 12 years ago, when the purchaser would fill out a web form, submit it, and it would result in an email to a clerk, who would turn around and type the order into the existing order management system. That’s not what ecommerce sites do today, however. And with the continued cost pressure on IT, which I wrote about in my last blog post, it’s going to be hard to justify this “man in the middle” staffing, though many organizations will no doubt attempt to do so.

Winners: IT operations. Huh, this is a surprise, eh? Well, if the inevitable outcome of reduced friction (not to mention cost — that will be addressed in my next post on Cloud Computing Revolutions) is to increase demand for IT resources, someone is going to have to do the capacity planning. In a sense, the impact of cloud computing will be to shift the tasks for IT operations from tactical resource provisioning to strategic resource planning — with an emphasis on achieving the most efficient, lowest cost infrastructure possible. This is a far cry from the “your mess for less” outsourcing that has previously been the outcome of cost focus — this is about creating an automated, immediate search for the lowest cost, most available, most appropriate computing resources needed to fulfill a provisioning request. The most successful IT operations groups will be those that stop thinking about controlling allocating resource and begin thinking about locating resource.

As I noted at the start of this posting, revolutions results in winners and losers after a period of chaos, in which conflict and strife occur. Technology revolutions at the level of platform shift — think client/server or the move to web-based applications — cause enormous upheavals in IT organizations as they struggle to adopt the new technology and obtain its benefits. Cloud computing undoubtedly represents the latest platform shift and is causing three simultaneous revolutions. This piece has focused on the revolution in operations. My next two postings will focus on cost/payment patterns and application implementation.

Bernard Golden is CEO of consulting firm HyperStratus, which specializes in virtualization, cloud computing and related issues. He is also the author of “Virtualization for Dummies,” the best-selling book on virtualization to date.

Follow Bernard Golden on Twitter @bernardgolden. Follow everything from on Twitter @CIOonline