This page contains a Flash digital edition of a book.
D ATA CENTRES


Understanding data centre power efficiency Measuring data centre PUE By Ian Dixon, VP Data Centre Operations at Colt


Today’s modern business has transformed to run on its data, with every activity depending on it; from research and development, to manufacturing and accounts. The data centre needs to be able


Ian Dixon explains how data centre efficiency is now a board level issue…


to store the data and secure it, whilst providing instant access on a constant basis. In order to keep up with the growing demands from businesses, it needs to offer scalability, flexibility to deal with emerging technologies and rising energy costs, all whilst trying to run in an efficient and environmentally friendly manner. Traditional data centres were


planned around a 10 to 15 year life cycle, however it’s almost impossible to balance long term planning, ever- changing business requirements and technology innovation. With business and technology


constantly changing even the greatest technology engaged business case cannot predict the longer term. Trying to plan anything beyond two years is a challenge, so ‘guesstimating’ computing requirements fifteen years in advance is nigh on impossible. The data centre in particular has


moved forward dramatically in the past ten years in terms of what is considered to be best practice. Previously, each data centre was built differently and operators had their tools and processes to run them.


However, today there is a whole host


of information available about how data centres should be run, how to improve efficiency and the best tools to use to improve and automate processes. Nevertheless, there still remains no single defined set of best practice standards. Typical of a developing industry,


the data centre sector is notorious for deploying measurements which seemed good at the time but moved out of fashion as the real issues surfaced. When the cost of energy rose, the


power of the data centre became a key factor for measurement, along with an increased focus on energy efficiency and the social responsibilities associated with running large data centre estates. This caused the operational


management team right up to the board level to prioritise improving efficiencies as a way to make measurable differences as well as an impact to the company’s bottom line.


The PUE rating Measuring Power Usage Efficiency (PUE) is one of those measurements and has become the de facto standard for measuring energy efficiency, environmental impact, and cost of running a modern data centre. As efficiency became synonymous to every data centre conversation, PUE


became the metric of choice – the smaller the better. At Colt, we were well aware of


the benefits of reducing power consumption and a couple of years ago we decided to regulate our approach to improving efficiency. We worked out a standard set of


guidelines for our operations crew to systematically ensure that from simple to more complex solutions, we were maximising efficiency in every data centre we have. As a result we reduced our annual


power bill by more than €4 million. With that in mind we strive to help


others benefit from improving their PUE, which is why we have compiled a list of 12 steps to improve PUE.


12 steps Measure - in order to assess any efficiency improvement, you need to be in a position to benchmark your existing energy usage against a comparable timeframe. Starting to measure, record and track power use on a regular basis is the first step to a more efficient data centre.


Regulate airflow – prevent hot and cold mixing - The need to regulate temperature can be one of the most draining processes on energy supply, and have a substantial effect on PUE. A data centre is essentially a structure that manages the flow of cold air in, and extracts hot air. With servers pumping hot air out and cold air in, if not managed correctly mixing will occur, so controlling airflow and limiting mixing; and the various steps to do so should be a priority.


Align hot and cold aisles - ideally, all servers within the racks and rows should face the same direction. These rows can then be arranged into hot and cold aisles with the front of the servers facing one way and the back the other; to allow for cool air aisles to be created blowing colder air to the front of the servers, and hot air aisles for the hot air to flow from the back.


Empty space = wasted heat 22 NETCOMMS europe Volume IV Issue 3 2014


Reduce airflow leaks - install blanking plates to fill the gaps where


www.netcommseurope.com


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44