Recent research is adding impetus to the notion that the cloud will abolish the data center as we know it. And in fact, there is a kernel of truth to this, although the change is not going to be as complete as some cloud enthusiasts would have us believe.
As I noted in a previous post, companies like IDC are projecting a steady increase in new data center construction over the next few years, followed by a drop-off as workloads are moved from internal infrastructure to external resources that are either cloud-based or provided in standard colocation fashion. This is not necessarily a death-blow to the data industry, however, given that scale-out cloud architectures will still require a fair amount of hardware and software – although most of it based on low-cost, commodity designs – so that the amount of actual data center square footage is likely to increase well into the next decade.
So clearly, the enterprise data center will not vanish, but it will change. For one thing, smaller facilities and “data closets” are likely to fade away, leaving larger, scale-out data centers in their place. As well, look for increased density and modular designs rather than today’s rack configurations, says Dell’s Ashley Gorakhpurwalla, with much of today’s hardware-driven functionality replaced by advanced software-defined architectures. The good news, though, is that much, if not all, of this new infrastructure will integrate seamlessly with legacy systems, giving the enterprise crucial breathing room when it comes to building next-generation data environments.
Clearly, though, much of the enterprise workload is going to make its way onto third-party cloud infrastructure, and this presents a problem when it comes to management and governance. This is why market research is also leaning toward a boom in data center automation for the rest of the decade. Research and Markets, for one, predicts annual growth of more than 18 percent until 2018 as organizations look to remove the human component from routine tasks like scheduling, configuration management, patching and updating. The idea is not to create a fully self-operating infrastructure (SkyNet, anyone?) but to increase the speed and accuracy to which data loads can be matched to available resources.
Extending management tools into the cloud is difficult enough, but many organizations are still struggling with the fact that they don’t even have visibility into public infrastructure, leaving them completely in the dark as to what is happening out there until service suddenly vanishes. Providers like Microsoft, however, are hoping to rectify this situation with the HDInsight service that seeks to unify data center and cloud visibility to provide real-time operational intelligence to those overseeing distributed data architectures. The service is geared toward optimizing infrastructure for Big Data applications, allowing for rapid configuration of Hadoop clusters and streamlined migration of business intelligence and other analytics workloads.
All of this activity points to continued reliance on internal data infrastructure for some time to come. Will there be some organizations that opt to place their entire data footprint on the public cloud? Undoubtedly—particularly among start-ups that need to avoid substantial overhead until revenues start coming in. But if past is prologue, many small organizations will begin to build their own data facilities to handle critical functions once business lines are established.
The local data center will be a different animal once the cloud era truly gets under way. It will be smaller, leaner and more closely focused on key applications that drive worker and business productivity. And best of all, it will cost less to build and operate with very little of the dead weight that inhabits today’s infrastructure.
Arthur Cole writes about infrastructure for IT Business Edge. Cole has been covering the high-tech media and computing industries for more than 20 years, having served as editor of TV Technology, Video Technology News, Internet News and Multimedia Weekly. His contributions have appeared in Communications Today and Enterprise Networking Planet and as web content for numerous high-tech clients like TwinStrata, Carpathia and NetMagic.