Enterprise Architecture in the Software-Defined Data Center

Arthur Cole
Slide Show

Eight Critical Forces Shaping Data Center Strategy

I’ve said many times in the past that physical infrastructure is and will remain a crucial component of the data environment. After all, software isn’t much good without a solid hardware foundation. But as virtualization and software-defined architectures continue to work their way into the enterprise, it is also clear that the majority of enterprise management activity will shift to these higher level architectures.

Hardware, in other words, will be important, but boring. And that poses some interesting questions as to how data environments are to be built and managed, particularly in the way the burgeoning field of enterprise architecture (EA) will come to supplant many traditional IT roles.

Forbes’ Howard Baldwin offered some interesting insight into this topic recently, noting that enterprise architecture is not so much about optimizing systems for current loads, but laying the foundation for an  environment that can be redesigned and repurposed without forcing an infrastructure rebuild. That, of course, is a tall order at the moment, but the hope is that with the triple combination of server virtualization and software-defined storage and networking, the advent of the fully dynamic virtual data center is finally upon us.


But this kind of functionality will not arise merely as a by-product of virtualization. Rather, it will have to be designed from the ground up, and the simple fact is that no one has a clear idea yet of exactly how to do it. As ZDnet’s Joe McKendrick pointed out recently, requirements for the enterprise architect (EA) are all over the map depending on whether the organization is focused on e-commerce, commodities trading, government contracting or any number of vertical industries. At best, we can guess that the EA will be responsible for a wide range of daily management and governance tasks, as well as long-term strategic planning and policy development, not to mention conflict resolution as individual business units seek to tailor the architecture to their own narrow ends.

Of course, logic would dictate that control of the enterprise architecture should rest with IT, given its knowledge of the inner workings of the infrastructure, platforms, applications and everything else that supports data activity. But recent research from the Vlerick Business School in Belgium suggests otherwise. Enterprise architecture, after all, is a means to enhance business functionality, so it only makes sense that business units should be in a position to determine what they need from the data environment, rather than conform to what an IT-centric architecture provides. In that vein, the school has launched a new Centre for Excellence in Enterprise Architecture Management as a means to empower business leaders with the knowledge they need to navigate the new digital paradigm.

To long-time IT executives, no doubt, this is the equivalent of barbarians at the gate. How are non-technical people supposed to manage an as yet largely unknown entity when even leading experts can’t seem to agree on exactly how it is to be built? For instance, should the architecture be flat as tech evangelist Tom Graves recommends, or should it be layered on business, technology and user planes as preferred by the UK’s Adrian Grigoriu. On the one hand, a flattened architecture may provide more balanced functionality for multiple business units, while the layered approach may lend itself to greater centralized management and control. Again, the central question seems to be: Whose needs are paramount, IT’s or the business units’?

With many enterprises still working out how to effectively implement basic server virtualization, the details of this still-theoretical enterprise architecture may seem like something for the back burner. But the future will be here before you know it, and the fact is that decisions made today, even those regarding lowly physical infrastructure, can have a significant impact on the kinds of architectures available to you tomorrow.

At this point, we can’t work out all the kinks, but it’s never too early to think about where you want to be when architecture becomes the new driver for productivity.



Add Comment      Leave a comment on this blog post
Jan 13, 2014 8:07 AM Michael Bushong Michael Bushong  says:
I was struck by the opening salvo suggesting hardware becomes boring. I am not sure I fully agree there. I think that edge policy will definitely move off the boxes. Building all that functionality into the hardware platforms is likely unnecessary, and things like SDN and network virtualization will make it possible to separate the two. What you are left with, though, doesn't have to be dumb, static pipes. If anything, once you free yourself from the pain of managing on a device-by-device basis, you will want your underlying transport to be more flexible and, dare I say, responsive to the applications. There are opportunities to marry photonic switching with SDN that create dynamic bandwidth capabilities. Imagine an application requires a bunch of capacity for a short period. You could allocate that capacity and then return it to the network when the app is done. The hardware works in cahoots with the software intelligence. I don't know that the future is just the same networks we have today but done more cheaply. That seems an awfully low bar to aim for. -Mike Bushong (@mbushong) Plexxi Reply

Post a comment

 

 

 

 


(Maximum characters: 1200). You have 1200 characters left.

 

 

Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.


 
Resource centers

Business Intelligence

Business performance information for strategic and operational decision-making

SOA

SOA uses interoperable services grouped around business processes to ease data integration

Data Warehousing

Data warehousing helps companies make sense of their operational data