Prefacing his keynote presentation “Driving the next 80% improvement in data center performance,” which Kevin Brown delivered at DatacenterDynamics Zettastructure in London, the SVP of Innovation & CTO for Schneider Electric’s IT Division described a new way of segmenting the types of physical infrastructure into three very different sorts of data centers.
Categorizing data center environments
Following his presentation, I asked Kevin to dig a little deeper into what he’d said, and he explained that his categorization emerges from the vocabulary which is being used today to talk about the edge. In effect, the way it’s being classified allows us to simplify the market into three tiers, consisting of centralized cloud data centers, regional edge data centers, and then local edge data centers (the latter being yesterday’s networking closet and server rooms).
According to Gartner, the edge is essential to the computing requirements for emerging applications. It promises near real-time insights and facilitates localized actions by moving data processing closer to its use or user. Gartner predicts that edge growth will be significant: “Currently around 10% of enterprise-generated data is created and processed outside a traditional centralized data center or cloud. By 2022… this figure will reach 50%.”
The challenges of burgeoning edge data centers
In an edge driven future, all forms of both centralized and distributed IT environments will be classified as data centers. The outcome of this is that within some enterprises, the IT team may be tasked to manage hundreds, if not thousands of data centers as their networks evolve.
With that in mind, I asked Kevin what he felt were the key challenges presented by the proliferation of edge data center environments. He said that if we looked back a decade, the teams that managed data centers in any organization probably had no more than one or two facilities to manage and control. Such facilities were well identified and well invested with high security, great management and full control over what happened.
But things are starting to look very different today. Teams are looking after a much larger number of distributed IT environments which may have been deployed in less than ideal situations – for example, in office space or corridors, even under desks and in stairwells. Nonetheless, these facilities may be critical to ensuring vital processing, storage and/ or connectivity for the workforce. The challenges therefore include not only how to power, cool and protect the equipment, but how to manage the system effectively and ensure its physical security.
Kevin told me that Schneider Electric’s view is that many data center management tools that have been used up until now really are going to be inadequate in the edge computing future. This is because once you get into this complex, hybrid data center environment which may comprise multiple local edge, regional edge and cloud data centers, a different approach to management is demanded.
Managing complex distributed IT environments
For Schneider Electric, enabling data center management tools in the cloud will help provide the ability for customers to have the control, resiliency, and security that they need across this entire hybrid environment. Kevin said that Schneider Electric’s EcoStruxure fills this space today because it has been developed as an integrated, IoT-based solution. Watch my interview with Kevin to explore how the industry can make the next 80% increase in data center efficiency.