In modern data center design, the most overlooked elements are power utilization, consumption and monitoring. Even when a data center is built out, thinking about power usage must always be at the forefront of any IT administrator’s mind. Making the right decisions can improve server health and data center efficiency. On the other hand, overlooking simple things, like monitoring data center power consumption, can have detrimental and costly effects on the environment. In this tip, you’ll learn some strategies to manage data center power.

Creating a power-conscious data center
The idea is simple: You can’t effectively manage what you can’t monitor. The only way to run a properly optimized data center is to have a solid monitoring platform designed for your environment. But every data center is unique, and power requirements will vary from machine to machine and site to site. So it’s important to research what you have now, where you want to be five to 10 years from now, and how technology purchases will affect your power usage–both long and short term.

Monitoring power consumption. To gain a solid, overall view of data center power consumption, power utilization should be monitored at several points. These include the uninterruptible power supply (UPS), the room power distribution unit (PDU) and within the actual rack environment. An important task of power management is to understand and calculate the power usage effectiveness

(PUE) of a data center. To do so, measurements should be taken at the UPS, which will then provide a base measure of data center power consumption. The measurements will then be used to calculate the PUE and identify energy usage trends. By monitoring the PDU, administrators can prevent certain overload conditions at the PDU level. By effectively monitoring aggregate PDUs, admins can also help ensure that power is evenly distributed across the facility. Arguably, the best view of data center power consumption will come from the power distribution units located within the racks. Modern rack-mounted PDUs feature integrated monitoring and control capabilities to enable continuous power monitoring. These PDUs are further configurable to send alerts and updates to a centralized management agent for data analysis.

Since rack and data center power consumption numbers vary based on the specific equipment within the infrastructure as well as the respective loads, each individual rack should be fitted with its own PDU. For some more advanced environments, power engineers recommend using two PDUs for dual-bus environments. Using two PDUs helps in two ways: It enables the capability of monitoring power consumption to the rack PDU and provides more overload-protected receptacle groups.

With modern power management systems, administrators are able to provide PDU, branch-level and receptacle-level monitoring of volts, kilowatts, amps and kilowatts per hour. Energy architects are then able to use the most direct measure of data center power consumption available. This information can be used at the data center management level and supports both higher data center efficiency and availability. Remember, monitoring power is crucial, but what about wasted resources? A solid, rack PDU setup not only provides effective power management, but it can also be used to support chargeback of IT services and identify stranded capacity. Modern rack and room PDUs can also enable individual receptacles to be controlled remotely to prevent the addition of new or uncontrolled devices that, if not monitored, could create a dangerous overload condition.

Monitoring energy efficiency. IT administrators know that energy costs can quickly consume a large proportion of a data center’s operating costs. Unfortunately, given the facts, many facilities still lack solid energy monitoring capabilities. The answer came from the industry in the form of centralized monitoring systems. Centralized monitoring systems, like the kinds from Emerson Electric Co. or Liebert, are available today that operate across the existing IT network or across a dedicated network. When creating a dedicated power monitoring environment, the suggested best practice is simple:

• Sites between 2000 and 3000 square feet generally choose to use the existing network infrastructure, rather than set up a separate network.
• Larger facilities with bigger data center requirements will benefit from a dedicated network. This can provide the ability to integrate with building automation and management systems, as well as manage multiple corporate sites.

Managing data center power does not have to be a manual process. Automating the collection process and then analyzing the data from the respective UPS and PDU monitoring systems can help reduce energy consumption by tracking efficiency. The other major benefit of having an automated system is increased IT productivity. Energy efficiency monitoring allows administrators to track total data center consumption, automatically calculate and analyze the PUE and further optimize the use of alternative energy sources. UPS devices carry a wealth of information. Monitoring systems can track UPS power output and determine when these units are running at peak efficiency. Using modern, power-conscious systems, even monitoring power at the panel is beneficial. Choose a system that checks power usage outside of the normal server and rack parameters, and looks at power consumption by non-IT systems, such as lighting and generators, to ensure the absolute best use of those systems.

Power monitoring tips and benefits
There are numerous benefits to monitoring and managing power usage:

• Increased profitability with lowered energy and operating costs: Even a small reduction in energy consumption can deliver substantial cost savings over time. 
• Rebates and awards: Many organizations, local and regional governments, as well as energy vendors, offer energy rebates and awards that can help offset the cost of implementing newer, more energy-efficient technologies. Energy-conscious organizations, such as Energy Star, The Green Grid, and the Uptime Institute offer awards and certifications for energy efficiency. A number of U.S. states offer rebates for energy-efficient technologies and practices. You can find a list of rebates by state at the Database of State Incentives for Renewables and Effeciency website.
• Accurate chargeback: Monitoring power provides accurate data on usage, making it easier to compute charges that can then be used to chargeback computing users more fairly.
• Reduced emissions: Efficiently managing power usage in the data center not only cuts your own costs and carbon footprint, but also lowers overall emissions.

Remember, the little things can make all the difference when it comes to successful power consumption monitoring. Below is a short list of often-forgotten, power-critical components.

1. Keep an eye on your server room temperatures: By installing network temperature sensors across a data center (within racks, exhausts, hot/cold isles, etc.) administrators can help ensure that their equipment is running within the ASHRAE recommended range (64.4 to 80.6 degrees Fahrenheit). With cooling in check, data center power consumption can also be normalized. This helps keep the servers running properly and keeps power costs down.
2. Manage your alerts and alarms: Automated systems can help keep power usage in check and lead to a healthier environment. Create alerts and alarms so that administrators can take a proactive approach to a very important data center function--power management. By knowing a problem area before an actual issue arises, admins can take necessary actions to offload power or redistribute it as required.
3. Keep your batteries healthy: The best way to help prevent data loss or server damage is to have a properly configured and healthy backup battery environment. A best practice is to implement a UPS or battery monitoring system that is able to connect to and track the health of each battery within a string. The most effective battery monitoring systems will continuously track all battery parameters, including internal resistance, which means using a DC test current to ensure measurement accuracy and repeatability. A good routine of preventive maintenance, battery replacement programs and UPS monitoring can help keep data centers and battery systems up and healthy, which will pay dividends when power disruptions occur.

About the expert: Bill Kleyman, MBA, MISM, is an avid technologist with experience in network infrastructure management. His engineering work includes large virtualization deployments as well as business network design and implementation. Currently, he is a virtualization architect at MTM Technologies Inc., a solutions provider based in Stamford, Conn.

Disclaimer: Our Tips Exchange is a forum for you to share technical advice and expertise with your peers and to learn from other enterprise IT professionals. TechTarget provides the infrastructure to facilitate this sharing of information. However, we cannot guarantee the accuracy or validity of the material submitted. You agree that your use of the Ask The Expert services and your reliance on any questions, answers, information or other materials received through this Web site is at your own risk.