Are PUE Metrics Skewed?
On the subject of data center energy costs, the debate is far reaching. Whether you’re established or a start-up, the topic of energy consumption comes up, it comes up even – or is that especially – if you’re the government. Energy Star, small foot print; catch phrases abound in the discussion. One thing that sometimes gets over looked, perhaps on purpose, is how PUE, or Power Usage Effectiveness, is measured and then reported. Varying factors can play havoc with PUE; weather, season, Tier classification, and if a data center’s PUE is only measured during one month out of twelve then we can see a disparity – or worse, an exploitable gap. In typical tech industry savvy a group of people got together on LinkedIn and discussed solutions and some of the other hurdles being faced.
The bottom line? The bottom line. Building smarter and more efficient models in the vein of Google means a better bottom line. Whether you calculate PUE with an annual span or a month to month average, you’re looking to make sure you stay within a certain range. When you fall out of that range how do you determine the cause? What do you do to ensure efficiency of energy in your data center? Feel free to comment below with some strategies that have or haven’t worked for you.
Check out the link to read more on the LinkedIn meeting and an in depth look at the proposed solutions:
Comments are closed