As a marketing wise-acre, I could joke that "cloud" is a marketer's dream -- it's so vaporous, you can make it mean anything you want, right? Actually, most us here have a pretty precise idea of what it is, and what it might mean.
By the way, we're in the middle of a series of posts here.
Just to bring you up to speed, I've introduced this series, written about the growing need for information governance, identified information risk management as the new frontier in security, pointed to the unmet needs of knowledge workers as a crisis in the making, described how the changing nature of applications will change IT, and positioned virtualization as creating the potential for "frictionless" IT.
Now we're ready to connect a few of the dots, and take on the concept of cloud computing.
And, before long, I'm guessing it'll be a very interesing discussion to most of us in the industry.
The Multiple Meanings Of Cloud
At its simplest level, "cloud" is any information service I get over a wire. I don't care much where it's done, or how it's done. A useful simplification, but there's more to the discussion.
The premise is simple: if a cloud is focused on a specific task (e.g. search, backup, collaboration, security monitoring, etc.) it can be built extremely cost-efficiently, due to scale and optimization effects. At the same time, since it's purpose-built, it can probably do a better job at a given task than you can do yourself.
Put differently, general-purpose IT is hard to reconcile with the concept of cloud; it's more productive to think in terms of optimized services delivered by a specialist.
If you're in IT, that brings up two interesting questions.
The first one is -- how much of this stuff do I want to do myself, and what pieces would I be willing to trust someone else to do on my behalf?
The second, less-asked question is -- do I do any large-scale tasks internally that might be useful to think of in terms of an "internal cloud?"
But, before you answer these questions, let's connect a few of the dots that might point to an expanded frame.
Application As Service Compositions
In a previous post, we discussed how the fundamental notion of an application was changing: tomorrow's application is best thought of as services that are composed to implement business processes.
In a world of decomposed services, how many of those make sense to run behind the firewall, and how many might make sense to contract externally for?
If you're a small, Web 2.0-ish type business, the answer is "almost all of them", hence the popularity of Amazon's environment.
Disruption has this nasty habit of starting at the bottom, and working its way up.
We also talked a bit about how the infrastructure -- and the management of this infrastructure -- would decidedly be different. How many of these infrastructure and related management services would best be done by a specialized provider?
Or, put differently, how much of your precious IT resource will you want to put behind things like backup, web serving, email, records retention, collaboration, security monitoring, service delivery monitoring, and so on?
Whatever the answers are today, they're likely to change in the future.
Virtual IT
In another post, I discussed how virtualization essentially takes the "friction" out of much of IT. Resources are liquid, they can go where they need to go, when they need to. Including to another location entirely.
Which, of course, brings up the notion of a "virtual machine cloud" that you can easily move things to (or from) when you need extra capacity, server images, virtual desktops, geographic separation, etc.
By taking the friction out, a whole new set of cloud providers can be created, giving IT even more options.
Supporting The Knowledge Worker
And, finally, in still another post, I lamented on how existing IT models tend to underserve an important value-generator in our organizations -- the knowledge worker.
We're already starting to see growth in "clouds" that are targeting these needs; perhaps the best example is Cisco's WebEx, which *can't* be implemented behind the firewall; you have to consume it as a service on their cloud.
What about social productivity software? Document-oriented collaboration? Maybe those things make sense when considering a single enterprise, but using a trusted cloud makes even more sense when considering interactions between companies or other organizations.
Which participant in a two, three or multi-way collaboration is going to open up their firewall to others? It's not going to happen, at least not easily, hence the attractiveness of a third party.
New Information Types, New Consumption Models
And, finally, way back in the introduction to these series, I pointed out that the vast majority of new information was becoming unstructured content -- generated by individuals, consumed by individuals, who are increasingly becoming mobile.
If you think of the idealized technology that could capture this information anywhere, and deliver it anywhere, you tend to think less in terms of a centralized, massive data center, and more in terms of multiple nodes that move this information from where it lives to where it's needed, dynamically and automatically.
A content cloud, as an example.
Is There A Cloud In Your Future?
Well, if you've ever used Google, that day is here already, isn't it?
The question is -- how will our organizations leverage the new clouds that will undoubtedly be built in the near future?
And, which ones will we want to build for ourselves?
Comments