DCN May 2016 - Page 25

design & facilities management considering the tactics, approaches and security of creating a future proof solution. Typically, the environment can significantly influence the approach and ultimately the ongoing management of the data centre. The impact of the nature of the space (has it low ceilings, does it have a raised floor etc?) How big is the space? What restrictions are involved? The list goes on. Plus aside from knowing the technical specifications of your UPS, the need to know your building’s structural limits is essential. Protective UPS equipment is often heavy and there is no going back if it all falls through that mezzanine floor! Functionality vs. cost management needs to be in balance with the need for protection. But protect it from what? What are these crucial elements that need to be woven together? Simply put, consideration needs to be given to three principle aspects; the environment, the power and the temperature. Good system design should address each of these, while From a power and energy management perspective, it is imperative to get the right equipment to achieve the support required, while achieving a balance on functionality versus cost. The initial investment will pay you back in the future and be worth every penny spent. Consideration should be given to the type of power distribution unit (PDU) or mains distribution unit (MDU) required , to accurately distribute electric power to the racks of computers and networking equipment. Effective design also extends to exploring ways to increasing efficiency and reducing costs, with many modern UPS now being able to operate in a range of modes, such as ‘economy’, when loads are light or inactive. Moving on to cooling and maintaining temperature, a data centre’s requirements are extensive. With all that equipment, the amount of heat generated is likely to be significant and will need to be addressed to avoid potential overheating and fire. Every aspect should be considered from in-row, fresh air or portable cooling to specific computer room air cooling (CRAC) to ensure cold air flows directly through the racks, picking up the heat as it goes, before exiting. Monitoring and maintenance There is something to be said for the upfront financial outlay for ongoing monitoring and maintenance, which is likely in the long term to save money. This should give a detailed and realistic insight into how your server and the environment are operating, in real time without necessarily having to be on site. Ideally, including trigger alerts via email, SMS, cloud based reporting or having hardware plugins to keep people informed about environmental changes on a 24/7 basis. The ability to monitor bandwidth use, as well as energy, storage, physical rack space will minimise downtime and improve efficiency through optimised power generation, transmission and generation. Unfortunately the realities are it all costs money. But a holistic approach to the design, where each element is considered in isolation, but intrinsically interlinked, managed and monitored will deliver long term gains. All focused on offering the lowest total cost of ownership, while also delivering long term future flexibility. 25