Staying chilled: data centre power & cooling challenges

There’s no doubt that the challenges of cooling a data centre are tricky. Global specialist in energy management, Schneider Electric, advises focusing on the following four core challenges to get a head start in keeping a company’s data centre power and cooling infrastructure in optimal shape. These are: increased power densities, poor power usage effectiveness (PUE), increased risk of downtime and unnecessary levels of redundancy.

Areas of high-density pose heat removal challenges, and in cloud environments, virtualised servers are often grouped in ways that can create “hot spots”. Schneider Electric advises its customers that the preferred solution is to move higher density equipment to specific racks or rows, separate from lower density equipment, as this high density ‘pod’ ensures more predictability, and maximum density per rack.

Virtualisaton can result in a significant decrease in server energy consumption. However, if the power and cooling infrastructure is left in its pre-virtualisation state, then PUE worsens. Scalable power and cooling solutions, available from Schneider Electric, can significantly reduce these inefficiencies by “rightsizing” to the new load.

The energy management specialist is also fully cognisant that loads within cloud data centres can vary in both time and location. That is, a highly virtualised cloud-based data centre will tend to experience larger load swings. For this reason, it advocates predictive management tools as a safe way to maintain an efficient, lean data centre in a cloud environment. For instance, tools such as Schneider Electric StruxureWare for Data Centres suite work in real time with Virtual Machine (VM) managers to assure that power and cooling capacity can efficiently accommodate the dynamic loads.

Lastly, cloud data centres are operated with a high level of IT fault-tolerance and, as stated by Schneider Electric, “this may reduce the need for power and cooling infrastructure redundancy”. Accordingly, matching of physical infrastructure redundancy to the fault-tolerant nature of a virtualised cloud environment is another form of the rightsizing, and can save significant costs.