When assessing colocation providers, it’s easy for companies to get caught up in comparing connectivity options and other data center services. While these are clearly important, a good data center strategy should also take a good look at how well a facility’s infrastructure and operations have been optimized. An organization that fails to take optimization into account will very often wind up paying far more than they expected for colocation services, with the facility passing along the cost of its inefficient infrastructure along to customers. That’s why optimization should be among the most important questions a company asks about a data center.
Data center optimization refers to any series of processes and policies that are implemented to increase the efficiency of data center operations without sacrificing functionality or performance. Generally speaking, an optimization strategy’s effectiveness is measured by the facility’s power usage effectiveness (PUE) score, which compares the total power requirements of computing equipment against the data center’s total power consumption. When looking at two comparable data centers, the one with the lower PUE score has generally implemented various optimization strategies to reduce the power consumption of its IT equipment.
This has important implications for colocation customers. Facilities with low PUE scores deliver better value, making it possible to deploy more powerful computing and network solutions at a far more competitive price point. Less efficient facilities pass their high energy and cooling costs on to their customers, which severely limits flexibility. A low PUE score also provides a predictable IT environment that allows colocation customers to scale operations cost-effectively.
For all of these reasons, it’s critical that an organization’s data center strategy prioritizes optimization.
Any optimization strategy must begin with data. Collecting information about network utilization, server power consumption, and cooling performance establishes important baselines for measuring the efficiency of a data center’s infrastructure. By implementing a data center infrastructure management platform like vXchnge’s award-winning in\site, data center managers can gain unprecedented visibility into how their facility is using power on a daily basis. These programs monitor features like server utilization, network traffic, and cooling usage, revealing where and when data centers are using the most energy. Strategies like shifting compute workloads, powering down servers when they’re not needed, and cycling up resources during high demand periods can greatly improve energy efficiency without compromising performance.
Organizations often make the mistake of purchasing computing equipment that only meets their existing needs. Unfortunately, when their business grows, those assets won’t be able to keep pace with their new computing and network demands. The quick solution in such a scenario is to provision additional equipment to fill the gap, but this is rarely good in terms of efficiency. Running a single, high-density server cabinet, for example, is generally more efficient than running the same processing load across multiple low-density cabinets. Purchasing more powerful equipment that can accommodate future needs may carry higher up-front costs, but it will deliver better efficiency and performance in the long run.
The physical characteristics of a data environment are an important aspect of data center optimization. A poor server deployment could introduce latency into network systems, especially if a facility uses sloppy, unstructured cabling. The wrong deployment could also increase cooling costs if servers are unable to vent heat effectively. Workloads may not be allocated properly across computing resources, which will result in multiple servers running well below capacity, but still consuming as much power as servers running heavier workloads.
By optimizing their data center environment, colocation facilities can deliver a number of performance benefits to their customers. The ability to implement data center optimization strategies at scale is one of the key advantages colocation facilities have over private, on-premises data solutions.
Better Compute Usage: Whether it’s through server virtualization or hybrid cloud deployments, well-optimized data centers can manage computing workloads far more effectively. Low latency cross-connections to public cloud platforms allow colocation customers to scale their computing resources rapidly, leaning on the power of distant hyperscale data centers for big data analytics.
Better Cooling Efficiency: Cooling infrastructure represents about 40 percent of the total energy consumed by a data center. The problem is even worse in older, poorly optimized facilities, many of which try to overcome their inefficient design by adding more cooling capacity. Data centers that incorporate sophisticated DCIM tools to carefully monitor and regulate cooling resources, however, can significantly reduce their cooling costs.
Better Uptime Reliability: Well-optimized data centers also tend to be more reliable. Since they use energy more efficiently, there is less wear and tear on both IT and cooling equipment. Infrastructure redundancy means that no single point of failure can bring down a network, and the ability to monitor performance with DCIM platforms tools helps IT engineers troubleshoot problems more effectively to maintain high levels of network uptime.
Data center optimization is a key differentiator among colocation providers. By identifying facilities with the best data center strategy in terms of infrastructure and operations, colocation customers can take advantage of substantial energy savings and increased performance.