We have heard experts, technologists and IT professionals across orgnisations at all levels talk and talk and gush and gush (ad nauseam) about the benefits of virtualisation.
The cost efficiencies it immediately delivers and thereafter continues to yield, the flexibility it empowers the deploying organisation, the better to drive its strategic business growth…And we have not heard as much about the benefits, often great ones, that organisations glean or can derive from their virtualisation setups indirectly.
And in one of these benefit areas the global provider of power and cooling solutions for enterprise computing, with a special emphasis on data centres American Power Conversion (APC) figures very prominently as a leader. For an idea of how much APC has devoted to the task of helping its customers save the most, and perhaps gain even more, from their physical infrastructures after a virtualisation initiative, refer to the abstract below from an APC White Paper called “Virtualization: Optimized Power and Cooling to Maximize Benefits” (written by Suzanne Niles, Senior Research Analyst, APC Data Center Science Center).
Data centres are routinely and unknowingly missing a great portion of their entitlement from virtualization. Beyond virtualization’s undisputed IT benefits—from reduced rack footprint to disaster recovery—is the parallel story of a substantial benefit from optimizing the physical infrastructure that supports it.
3 Things You Must Know
There are three primary things to understand about virtualization as it relates to the data center’s power and cooling infrastructure:
- Power and cooling technology is available today to safeguard availability and meet the challenges of density and dynamics brought by virtualization.
- Power consumption will be less after virtualizing, as a result of the consolidation and physical reduction of the amount of IT equipment. With optimized power and cooling to minimize unused capacity, power consumption will typically be much less.
- Data center efficiency (DCiE) will go down after virtualizing, as the result of an increase in unused power and cooling capacity. With optimized power and cooling to minimize unused capacity, efficiency can be brought back to nearly pre-virtualization levels—sometimes higher, depending upon the nature of improvements of the cooling architecture.
Virtualization is bringing new extremes of power density and pace of change to the data center, increasing the demands on power and cooling infrastructure and naturally raising some concern about possible effects on availability. Fortunately, the key characteristic of virtualization—high density—is not new, and effective strategies for supporting it have had time to develop. While virtualization carries consolidated and dynamic computing to extraordinary new levels, the basic power and cooling requirements of virtualized computing are similar to those already introduced by high-density blade servers during the past decade. As a result, technologies are available today to meet the power, cooling, and management needs of a virtualized environment.
Physical consolidation from virtualizing will always reduce power consumption—directly, from the reduced server population and indirectly, from eliminating a portion of the power consumed by the power and cooling systems.
If power and cooling capacity is not optimized to match the new lower load, data centre efficiency will go down after virtualizing (even while power consumption is reduced), which reflects the additional overhead of idle power and cooling capacity at the new lower IT load. [There are available today] power and cooling systems that not only provide effective support for a virtualized environment, but also efficient support—significantly leveraging the IT-layer efficiency gains that are the signature benefit of virtualization. Properly designed physical infrastructure will not only provide solutions for the specific power and cooling demands of virtualization, but—especially if replacing room-based cooling systems—can raise bother power density capacity and overall data center efficiency significantly above what they were before virtualization.
The Challenges
Virtualization creates changes in the data center that present new challenges to power and cooling infrastructure, with implications to both effectiveness (how well it performs the job of safeguarding the IT load) and efficiency (how well it conserves power while performing that job). While an upgrade of power and cooling systems is not necessarily required to make virtualization “work,” the greatest benefits from virtualization will be realized with power and cooling that responds to these challenges.
The three major challenges that virtualization poses to physical infrastructure are dynamic high density, underloading of power/cooling systems, and the need for rack-level, real-time management of capacities (power, cooling, and physical space). These challenges are met by row-based cooling, scalable power and cooling, and capacity management tools, respectively. All three of these solutions are based on design principles that simultaneously resolve functional challenges, reduce power consumption, and increase efficiency.
Measured Efficiencies
Our comparisons of pre- and post-virtualization power consumption involved two concepts relatively new to data center cost analysis.
The first is fixed loss—the amount of power consumed by devices and systems regardless of load—which is responsible for the often surprising inefficiency of underloaded systems. The second is the distinction between energy consumption and energy efficiency, which can confuse a comparison of energy savings. Here is how these two concepts play in the virtualization discussion: Even without a parallel upgrade to power and cooling, virtualization will always lower the electric bill, but (1) not usually as much as might be expected because of the presence of fixed loss in power and cooling systems and (2) in spite of a reduction in power consumption by the data center, the DCiE is typically lower after virtualizing due to the inefficiency of underloaded power and cooling systems. This lowered efficiency indicates room for improvement in power and cooling systems—it is, in effect, a measure of the potential for extracting even more value per energy dollar.
Conclusion: Parallel Necessary
When virtualizing, a parallel upgrade of power and cooling infrastructure will optimize both architecture and operation in a number of ways that safeguard availability, enhance manageability, lower power consumption, and increase efficiency. Properly designed physical infrastructure will not only provide solutions for the specific needs of virtualization, but can also raise both power density capacity and data center efficiency significantly above what they were before virtualization.
