No matter what advances take place in enterprise infrastructure in the coming years, the largest cost center is likely to be storage. Even as infrastructure becomes more software defined, relentlessly increasing data volumes will require organizations to either buy or lease storage capacity in ever larger amounts.
The question, then, isn’t how to cut back on storage, as much as it is how to make more efficient use of available storage. As I’ve mentioned in earlier posts, even cloud infrastructure can start to cost dearly as time passes and data loads mount.
Already, we are starting to see a wide discrepancy in the amount of data being generated and the amount of storage being provisioned. Forbes’ John Webster noted recently that even though vendors like EMC are reporting yearly increases of about 10 percent, market analysts like IDC estimate data loads are climbing by about 70 percent per year. That can only mean one of two things: Either the enterprise industry is woefully underestimating the amount of storage it actually needs, or it is truly living by the credo “do more with less.” Again, though, the problem is that if today’s strategy is to simply offload storage requirements to the cloud, the ultimate bill for all that capacity will be monstrous.
New technologies, of course, are expected to put greater capacity in the field at lower cost, although it is questionable how effective they will be when loads are doubling every 18 months or so. Still, companies like Tegile Systems and Nimble Storage are pushing hybrid storage arrays that combine traditional spinning media and advanced Flash architectures in a bid to boost both performance and capacity without breaking the enterprise budget. At the same time, heavy use of compression, deduplication and other data management techniques are expected to bolster utilization and ultimately stretch the storage budget to its limit.
This is also part of the reason why venture capital is flowing toward the legions of storage start-ups that have recently come onto the scene. According to 451 Research, more than $1.2 billion was funneled to storage specialists in the past year, and that represents only the publicly disclosed capital. Apparently, the investor classes see storage as an area ripe for innovation and, like food and water, an essential commodity that will continue to produce healthy returns even if the killer storage app remains.
For established storage vendors, the saving grace just may be that enterprise customers are loathe to entrust such a critical function to a new-comer. And indeed, surveys seem to indicate that IT executives tend to favor storage infrastructure from trusted vendors who can supply integrated server and networking systems as well. The simple fact is that as data environments become more software defined, the need for close coordination between and among hardware and software systems is on the rise. And that means anyone who comes along with an innovative storage solution will first have to demonstrate that it can play nice with legacy infrastructure, both in-house and in the cloud.
Ultimately, it could very well turn out that decisions regarding storage, networking or any other IT function will become subsumed within the greater need to implement data infrastructure quickly and easily, most likely through a converged, modular approach. Infrastructure that scales in tandem, after all, will be better able to accommodate all user and application needs, not just provide a place to park data.