3 Ways Modern Data Centers Are Driving Energy Efficiency
Modern data center managers are constantly being asked to do more with less, and energy efficiency is no exception. Data center teams are facing increasing for IT resource capacity to store and process the data and applications required for big data, IoT, AI, and other transformative technologies. As a result, modern data center environments are more distributed and complex than ever before.
Maintaining a safe environment for their intelligent PDUs, UPSs, and other hardware equipment while also keeping costs low is a continuous struggle for today’s data center managers. Data Center Infrastructure Management (software) software can be a critical tool in helping you meet these objectives. Data center software can help to transform your current practices to reduce energy waste and decrease associated costs.
Here are three ways to optimize your data center energy consumption:
1. Look beyond Power Usage Effectiveness (PUE).
One of the most common ways of decreasing data center energy consumption is to optimize data center power usage effectiveness (PUE) to control energy costs. This approach makes sense when you consider that, based on Gartner research, power costs are expected to rise by at least 10 percent per year. These power costs are driven by cost per kilowatt-hour (kwh) increases and underlying demand. By 2021, power is likely to be roughly 15 percent of data center operating expenditure.
PUE is a ratio of the amount of power needed to drive and cool the data center versus the power draw from the IT equipment in the data center (in other words, a measure of your total facility energy versus your total IT equipment energy). A PUE of 1.0 is ideal, which would mean that you have a perfect data center. All of the power would come into the building, make it to the IT equipment without it being used for loads to cool or light up the room, and none lost in transmission to the IT equipment. Yet, many data center managers realize that PUE doesn’t provide a complete perspective of infrastructure energy efficiency. For example, it does not factor in the efficiency of your IT or networking equipment and it is only applicable to one site at a time.
You can’t manage what you can’t measure. Despite its limitations, PUE is still a prevalent metric and the industry standard for benchmarking data center energy efficiency. In fact, the average PUE in data centers hit a record low of 1.58 in 2018, according to Uptime Institute’s Eight Annual Data Center Survey.
However, modern data centers are also looking at other energy usage data to reduce their energy waste and to better understand how well they are using their data center space and power capacities. Some of these Key Performance Indicators (KPIs) are the same ones identified by the United States government as optimization metrics for assessing compliance with the federal Data Center Optimization Initiative (DCOI), including Energy Metering and Facility Utilization. DCIM software simplifies the process of deriving actionable insights by collecting and storing real-time data from the devices in your data center. The software then displays this data in easy-to-understand data center business intelligence dashboards and visual analytics. The software also enables you to manage your hardware and set thresholds and alerts on collected power and environmental data so you can keep an eye on your energy consumption.
2. Consider data center virtualization and consolidation.
In an effort to decrease the number of physical assets (and the expenses incurred from the space, power, and cooling capacities required to run them), modern data centers are looking to consolidate hardware and instead focus more on virtual assets. Whether this be through virtualization or leveraging a public/private cloud often depends on the organization’s data center asset management strategy.
Data center consolidation does not need to be on a small scale of certain equipment or even racks, although even minor efforts can decrease an organization’s IT footprint, energy consumption, and operating costs. Entire sites could be closed, with their workloads migrated to other physical locations or to the cloud. With fewer IT devices and without the costs of running and cooling these data center spaces, an organization’s energy consumption across data centers can be dramatically reduced. For example, server virtualization can improve overall utilization from roughly 10 percent to as much as over 50 percent, depending on the management system.
In addition to real-time power monitoring of your IT devices, DCIM software can also assist with asset tracking during your data center virtualization and/or consolidation. A comprehensive DCIM solution can track the location and details of each asset and expose performance indicators like CPU utilization/processor load, number of system processes, and memory usage. DCIM software can also display metrics around the number of hosts and virtual machines (VMs) per cluster, hosts per application and OS, and the number of physical versus virtual assets to help you ensure that you are fully utilizing your existing assets.
3. Bill back customers for energy usage.
Most organizations today have implemented green data center initiatives and other programs designed to drive energy-efficient behaviors. Compliance with these initiatives, however, is often more difficult to achieve.
Charging your internal and external customers for their data center energy consumption is an effective way to encourage them to model energy-efficient behaviors while also taking advantage of trusted data collected directly from your devices to reduce the potential for human error.
Data center software can take the data collected from your environments and display it in billback reports by customer, data center, or even rack. The ability to understand costs by customer, business unit, or application can help you identify power hogs and ensure that you are not under- or over-charging your customers.
Although reducing data center energy consumption continues to challenge modern data center managers, many are taking new activities to improve the efficiency of their own environments. Whether through real-time data collections, virtualization, consolidation, and/or customer billback programs, these efforts are leading to better utilization and decreased costs. When combined with DCIM software and other data center monitoring tools, these results can be even faster and easier to achieve.