Some organisations lack any form of data centre strategy. Others have an evolving set of practices built on past experience. Few have a specifically designed strategy that aligns with wider organisational objectives, consisting of a vision, a set of objectives or targets to achieve that vision, and a clear methodology. It’s natural to expect a set of credible, quantitative metrics at the heart of data centre strategy management, yet this continues to be a contentious issue rather than being the backbone of sound strategy development – particularly in data centre energy efficiency. The most frequently used energy metric in data centres is power usage effectiveness (PUE), which was derived by the Green Grid. A data centre with a high PUE is seen as inefficient, as a significant portion of energy usage goes into powering the facilities rather than the IT equipment. Conversely, a data centre with a low PUE uses most of its energy to power IT equipment. PUE is, however, somewhat crude and therefore not always well regarded by data centre architects. Arguments against PUE include that it’s only a snapshot in time, that it doesn’t take into account the utilisation of the data centre nor the efficiency of the IT equipment. As an example, using highly energy efficient IT equipment has the effect of increasing PUE, so simply attempting to optimise PUE could result in poor decision-making. The deficiencies of PUE have led to the creation of other metrics, such as carbon usage effectiveness (CUE) – another Green Grid metric to determine the relative carbon emissions of a data centre. In addition to these metrics, other common measurements used in data centre management are real energy consumption (in kWh), real energy cost (in monetary terms), or real carbon emissions (in tons of emission). Total cost of ownership is also used often to understand the full lifecycle cost of a data centre in monetary terms, comprising all components including design, build, operations, maintenance, energy supply, and disposal.
Related Articles -
Data, Data Centre, Data Centre Solutions,
|