Electronics Protection - September/October 2013 - (Page 6)

Feature Enabling Effective Thermal Management with DCIM Sev Onyshkevych, Chief Marketing Officer FieldView Solutions Data Center Infrastructure Management (DCIM) solutions help facilitate substantial opportunities for improving IT reliability, performance and cost effectiveness. No doubt, data centers are expensive to operate. A typical data center power bill runs into the millions of dollars, not to mention the cost of power and cooling equipment, IT hardware, maintenance, support staff and other operational expenses that increase the total cost of data center ownership to a staggering burden. A data center’s power consumption is an accepted cost of doing business, and these costs are constantly increasing, thanks to ever increasing use of applications, data and the evergrowing amount of equipment. However, strategies to curtail power costs often raise concerns that this process may impact IT services. Needless to say, a data center is a complex ecosystem of technology components where each element (the server, the storage device, the switch, the power distribution node, the environmental control unit) interacts in a variety of ways with every other element. Adding any new device directly affects the temperature, power draw and network performance of devices in the vicinity. Not to mention that neighboring devices will affect other components in a complex chain reaction that ripples across the data center. Getting an accurate sense of the power draw and temperature is extremely important if one is to try to reduce power costs without hurting the uptime (availability) of the IT infrastructure. One of the most common methods for estimating how efficiently one uses power is to calculate the data center’s Power Usage Effectiveness (PUE), which is accomplished by identifying the total data center facility power consumption and dividing that number by just the power consumed by IT equipment. The resulting PUE ratio indicates the proportion of energy consumed by the environment (the power distribution and cooling equipment) relative to the IT equipment it supports. Since the bulk of environmental power is often consumed by cooling equipment, a PUE can provide a fairly good, easily determined, indicator of thermal efficiencies. The average data center PUE is about 2.5, which means one-and-a-half times as much energy is consumed by the environmental and power distribution equipment than by the IT hardware supported. The goal of achieving thermal efficiency is to attain as close as possible to a PUE of 1.0. A realistically achievable goal would be a PUE of around 1.6, but this target can be higher or lower depending on factors such as the degree of redundancy, and on ambient temperature and humidity. However, depending on PUE as a thermal efficiency indicator on its own may be insufficient. Anomalies do arise; for instance, if servers are being consolidated or replaced with more efficient 6 ones, the PUE value may actually increase, since the IT equipment will consume less energy, while the power distribution and cooling chain may be less affected. In this case, even though the overall energy consumption costs have gone down, the PUE may not reflect that improvement. PUE is a much better indicator of improvements where data center managers have made adjustments to environmental conditions. In addition to PUE, one can garner more insight by holistically tracking power and cooling consumption, and using both real-time data and historic trends. To reduce costs and increase overall efficiency, a holistic view of the entire data center infrastructure is a must-have capability. Often times, this must have capability is provided by DCIM solutions that provide all the centralized monitoring and management processes necessary for enabling facilities and IT managers to optimize IT assets, while reducing CapEX and OpEx costs. An immediate benefit of a DCIM deployment is the improved thermal management of hot and cold areas. It’s no secret that heat-generating devices clustered in a confined area create hotspots. To address the hotspot issue, facility managers typically reduce data center temperatures, more than they need to, in an attempt to decrease these hotspot temperature zones. On the opposite end of the spectrum are cold spots. Chilled air can pool in one area of the data center if not properly circulated. This is often caused by cooling units not placed optimally, or not adjusted to the most efficient level, or by chilling areas where there is no IT equipment. If left to guesswork, data center managers usually resort to a fail-safe method of excessively lowering ambient temperatures. With each degree of reduced temperature, there’s a substantial increase in the cost of powering the cooling units. Identifying opportunities for reducing these cooling costs requires recording temperature readings from sensors and other live recording devices strategically distributed across the data center. Many DCIM solutions have the ability to automatically gather these readings and produce a thermographic image that clearly displays hotspots and other thermal challenges. With this real-time view, managers can now make informed decisions on how to optimally reconfigure the data center environment to minimize heat impacts. This reconfiguration may take many forms, from changing the physical environment to adjusting or relocating cooling equipment to relocating the heat sources (the servers, etc.) to cooling areas, to re-balancing the workload across a data center. A holistic view of the data center enables thermal and power improvements on both the Facilities side and the IT side of the house. Once improvements are made, regular monitoring of the thermal imaging provides IT support organizations with the assurance September/October 2013 www.ElectronicsProtectionMagazine.com http://www.ElectronicsProtectionMagazine.com

Table of Contents for the Digital Edition of Electronics Protection - September/October 2013

Geist Unveils Rapid Deployment Data Center Environmental Monitoring System
Enabling Effective Thermal Management with DCIM
Predicting Gasket Performance: SE Measurements with a TEM Cell to Study Gasket Reliability
Rack Containment 101
The Nine Core Elements of DCIM
Using Electronic Locking Solutions to Secure Enclosures and Meet Storage Compliance Needs
Hammond’s HJ Series Ticks All the Boxes
Rogers Introduces Poron SlimGrip Foam
Ferrite Suppressors Clear Interference
Fujipoly Thermal Sheets are a Cooling Influence on LED Lighting
GE Introduces TLE Series UPS Platform
Gore PolyVent XL Improves Reliability of Large Outdoor Enclosures
IMI Sensors Launches Linear Adjust Mechanical Vibration Switch
Industry News
Calendar of Events
Five Ways to Realize Server Room Profitability

Electronics Protection - September/October 2013