Most enterprises embrace the cloud because it provides a convenient means to scale resources and introduce a level of flexibility that has been lacking in traditional infrastructure. To many forward thinkers, however, this is the digital version of small ball – that without a full re-examination of data architectures, the enterprise will never realize the full potential of everything that emerging technology offers.
And it’s not just the cloud that is causing this massive re-think. Virtualization, mobile computing, unified communications and other developments are highlighting the many deficiencies that traditional data environments contain. Architecturally speaking, it’s like trying to fit a square enterprise into a round data universe.
ZapThink’s Jason Bloomberg, for example, argues in his new book, “The Agile Architecture Revolution,” that SOA, REST and other developments represent a true paradigm shift, not just in technology but in how that technology is used to accomplish goals. To date, most enterprise architectures have been integration-centric, driven largely by the need to incorporate disparate technologies into a cohesive whole. This was understandable given the technology of the day. But these days, the enterprise needs to adopt a more governance-centric approach, in which policy and procedures are the driving force for agility and flexibility in data environments.
If you think this is all ivory tower thinking with little or no consequence in the real world, think again, says tech guru Marcus Austin. Companies like music retailer HMV failed to adapt to new digital realities quickly enough and are now in receivership. Without the proper architectural framework, enterprises will find themselves behind the eight ball in short order given the speed at which modern technology is evolving. The most common solution is TOGAF, although a plethora of alternatives employ a wide variety of modeling strategies and methodologies.
Be wary, though, of data architectures, and the architects who propose them, if they promise complete digital nirvana, says Craig Martin, of consulting firm Enterprise Architects. In the first place, the sheer number of architecture models covering everything from business and processes to services and functionality can lead to architectural fragmentation if not implemented properly. Secondly, business architects and business analysts often work at cross purposes even though they are pursuing the same goal of producing predictable outcomes that meet objectives. Bridging these two disciplines, then, is crucial before we can even begin to envision data environments that approach the ideal.
Still, the modern enterprise has more immediate concerns that can and should be addressed by upgrades to architecture. One of them is the need for real-time performance, says Forrester’s Brian Hopkins. This can be achieved through techniques like highly granular predictive modeling, hub-and-spoke configurations as opposed to traditional layered designs, and instant analysis and change management. The cloud, mobile environments are other advancements are all geared toward real-time functionality. If the enterprise hopes to play this game, it needs to have both the proper equipment and the proper mindset.
On a positive note, the rise of virtual or software-defined infrastructure means that architecture is a lot more fungible than it was in the past, allowing data environments to be built, deployed and replaced much quicker and at much lower cost.
Going forward, a properly designed enterprise architecture will be a crucial component in the drive to keep up not only with today’s technology trends, but tomorrow’s as well.