No matter how virtual, cloudy or software-defined the data center becomes, it has to run on some kind of hardware somewhere. And that means someone has to pay the bills for operations, maintenance and upgrades.
But like everything else in the enterprise these days, automation is the name of the game, and as infrastructure becomes more consolidated, the data center infrastructure management (DCIM) platforms that oversee physical layer data, power and cooling resources are adapting to meet the new reality.
According to Navigant Research, the DCIM market is on track to top $4.5 billion by the end of the decade, a significant jump from today’s $700 million. Part of that growth will likely come from the need to clean up operations in legacy, static infrastructure, but a burgeoning demand exists to draw greater efficiency from up-and-coming virtual, consolidated systems. Doing more with less has become the mantra of the age, but that promise will remain unfulfilled if automation and optimization capabilities are limited to software-based abstractions.
Last week, for example, CommScope announced its new Data Center on Demand platform featuring prefab rack systems outfitted with the latest DCIM tools from recently acquired iTracs. That will likely include the Converged Physical Infrastructure Management 3.1 (CPIM) platform that is built around the company’s myAnalytics engine, which gathers and parses crucial operating data from across disparate infrastructure to enable improved change management and energy efficiency for highly dynamic data environments. Not only is it faster than traditional manual operations, but it reduces the chance of human error as well, improving reliability and availability of physical infrastructure.
Meanwhile, DCIM developer Rackwise has unveiled the new XTM platform, which offers single-line tracking of the entire power chain as a means to provide deep-dive analysis of consumption patterns and weaknesses. The system offers an interactive, single-pane UI that affords a diagrammatic view of power supplies—from the utility to the individual device—with simple point-and-click access to specific circuits and other points in the chain. As well, the platform provides rack and power provisioning tools, predictive change capabilities and end-to-end reporting.
Enterprise executives overseeing rapid change in the data center should take into consideration that future DCIM deployments need to address more than just the shift in physical make-up, but the new, rapid-fire style of resource provisioning and load management, according to HP’s Aaron Carmen. As he explains it to Dana Gardner at IT Director, static, monolithic infrastructure posed many management challenges, but they were nothing compared to the dynamic possibilities that modern virtual, converged platforms bring. Compounding the problem is the fact that most legacy architectures were governed by numerous management tools, each specializing in a key aspect of the data center, and these tools must somehow be brought under the yoke of an overarching DCIM platform if the enterprise hopes to gain the kind of efficiency and flexibility that software-based infrastructure promises.
Even vendors have to admit that the IT industry is entering a brave new world in which issues like management and optimization are still very much unsettled. The immediate concern for most organizations is getting software-based infrastructure to work at all—there should be plenty of time for fine-tuning later.
Still, that doesn’t mean DCIM should be pushed off entirely. As experience with today’s infrastructure shows, implementing a system-wide management stack only gets harder as more systems are brought online.