When I'm with customers, there's a lot of interest in the whole topic of "cloud" these days.
And there's also a lot of interest in NGDCs, or next-generation data centers.
But if you're interested in one, I'd argue you should be interested in the other -- as the two concepts may end up being connected in a way that's hard to ignore.
Cloud 101
Of course, there's no precise definition of a 'cloud' these days -- that'd be way too easy -- but a good generic definition is IT delivered as a standardized service.
Backing up corporate PCs, as an example, is attractive as a cloud offering: pay by the drink, everyone's requirements look pretty much the same.
More importantly, if you're a corporate IT group, that's the sort of task I'd describe as "hygiene": it's got to get done, but you don't get extra credit for doing it better than anyone else.
I am of the belief that -- over the next few years -- more and more "cloud" based services will force IT groups to reconsider what things they invest in doing themselves, and what things they hand over to someone else to do.
Note: I draw a distinction between cloud services and outsourcing -- with outsourcing, you're essentially contracting with someone to run your (presumably) customized environment; with cloud, you're running on a fairly standardized shared service.
We can debate the pros and cons -- but I think it's safe to say that there'll be many attractive cloud services in the future, and more than a few IT organizations will find themselves partaking.
NGDC 101
And, at the same time, there's a lot of interest in what data centers might look like in the same few years.
Common concepts include extensive use of virtualization to create dynamic pools of next-gen servers and storage, converged data center fabrics, model-based management that discovers, correlates and orchestrates resources, integrated security, and so on.
All of these technologies either exist today, or are on the visible horizon -- the real discussion is around how and when they'll be brought together in an integrated, coordinated fashion.
So, how are these discussions potentially related?
Two Ends Of The Same Spectrum?
One useful consideration is considering both ends of the spectrum. Cloud can be thought of as providing standard services that everyone pretty much needs.
As a result, it might make sense to assume that when you're envisioning your NGDC, there's an implied "cloud" of external services for you to leverage.
Logical candidates might be backup and recovery. Or security event management. Or monitoring of end-to-end service levels. Or perhaps long-term archiving of information. Or something else entirely.
Some of your current in-house application portfolio might be a candidate for a SaaS-like cloud model. CRM springs to mind, as does email. Or perhaps desktop productivity.
What's left in the data center after a serious consideration of cloud scenarios might be a distinctly smaller entity that's unique and core to your business.
We're talking customized applications, or perhaps things so uber-sensitive no one would seriously consider going outside the firewall.
Put differently, the effect of "cloud" can be thought of as Occam's Razor -- removing the extraneous non-value-added aspects of IT, and leaving the very core of IT differentiation.
Cloud As An Extension Of NGDC
A lot of the NGDC discussion has to do with hardware trends: what will servers and storage look like in this future world?
And -- at one level -- there's a viable scenario where this entire discussion might be largely irrelevant to many IT shops.
Why?
Once again -- virtualization changes everything.
Anything virtualized can easily be moved -- to a cloud, if desired. And it's not hard to imagine future "cloud" providers offering you a nice place to run your containerized virtual applications and desktops.
After all, the services required to run these virtual machines look pretty much the same across the marketplace -- making them a clear candidate for "cloud" providers to get into the game.
So, What's Left?
The answer depends, but try this on for size.
Imagine it's five years from now.
Look around your data center(s).
How many of the services you provide today could likely be provided by a cloud service?
Think backup, email, desktop productivity, content management, collaboration, file systems, archiving, aspects of management and monitoring, and maybe more.
Now, how many of your applications could run in a virtual container, and thus be a candidate to run somewhere else? And don't forget to include some hefty database-oriented applications on that list.
Consider what's left. I'd argue that -- in many cases -- the remainder is considerably smaller than before.
Of course, there are a variety of reasons why many IT shops may not go down this particular path -- historical precedents and resistance to change come to mind.
And there will always be uber-large shops where even this core remainder will be of considerable size and scale, necessitating a discussion of NGDC.
Cloud and NGDC
That's the problem with making connections -- once you make a connection between concepts, it's devilishly hard to pretend that the connection doesn't exist.
So, if I happen to get the opportunity to meet with you in the future, and we happen to get into a "cloud" or a "NGDC" discussion, and I stubbornly insist dragging one concept into the other -- at least you'll understand why.
Because I'm absolutely convinced that -- in the future -- you won't be able to discuss one without discussing the other.
Agreed. Cloud is really interesting. It will bring new dimensions and provides better servicing opportunities. But there is a catch. If the cloud exist beyond geographies, what will be laws and regulations that will regulate the data. Every country has its own laws and data stored in the cloud inside a country has to to follow it. This will be more interesting if the government tries to decipher the 'clouded' data, to validate the regulations followed.
Interesting!!! Doesnt it?
-- V
Posted by: Vijay | July 02, 2008 at 01:51 PM