Viridity supports thin power provisioning, energy cost reduction

By James E. Bagley

Viridity recently emerged from stealth after a two year period of developing strategies for addressing an increasingly critical component of data center operating expense, that is , power management. Every major piece of datacenter equipment carries a rated power consumption that exceeds the actual consumption of the components of the device, when running at peak, by as much as 100%. This differential leads to over provisioning of power resources and imbalances in power management which cause higher energy costs than necessary.

The analytics provided by Viridity will quickly determine this wasted energy volume through a combination of model-specific data, production measurements and statistical modeling. This “low hanging fruit” is just the first step in the process of analysis of energy use patterns within the data center in order to determine components drawing too much power, orphaned components that can be shut down, imbalances in cooling and airflow, and provide a level of infrastructure knowledge that is unavailable for current energy management products.

One of the basic tenets of quality improvement is the ability to create baseline measurements and continuous reporting in order to understand the causes and effects of variations in process. Intuitively, every data center manager will address power cost reduction from a standpoint of reduction of components due to virtualization, and the resulting reduction in power. While a simple concept, in execution it rarely works as planned. Often, after an application has migrated to a clean, new, energy-compliant server, the old, power-hungry system remains in service because it is too hard to determine if all applications using the resource have been weaned from it. By using resource discovery in context with energy measurement and monitoring at the distribution panel (not rack), and comparing the energy that should be used versus actual consumption, Viridity provides an entirely new dimension to data center information management.

Viridity, after an initial phase of close customer interaction during the start of an engagement, will be able to continue to add value at a level controlled by the customer’s continued participation in the input of parameters regarding its infrastructure. The delivery mechanism of Software as a Service will make it easy for customers to implement Viridity and gain increasing control of their infrastructure and energy consumption.

James E. Bagley is senior analyst for Storage Strategies NOW in Austin, Texas.