Data Center Forum. Switch N to Eaton. Forum focus: Distributed versus centralized power backup strategies

Transcription

1 Eaton's Global IT Newsletter Volume 6 Issue 1 Data Center Forum Forum focus: Distributed versus centralized power backup strategies Switch N to Eaton. IT Perspective Customer Voice Data Center Interactive

2 HE FINALLY UPGRADED OUR SERVER ROOM. WHAT WAS SO WRONG WITH LIVING IN THE PAST? Backup Power (UPS) 99% energy efficient with intuitive LCD display. Rack Power Distribution Manage power consumption. Power Management Software Optimized for virtualized environments. Say goodbye to weekends in the office. Discover our new power management solutions. It s time to reclaim your nights and enjoy your weekends: Things have changed. Eaton offers power management solutions you can count on from UPS products to airflow management systems, rackbased power management to power management software. Your desk toys will just have to keep missing you. switchon.eaton.com

3 Data Center Forum Letter from Brooke Lang: The Great Debate Brooke Lang Director, product management Most data center managers agree on the importance of using uninterruptible power systems to keep IT systems continuously operational. The best way to deploy those UPSs, however, is a more controversial topic. Generally speaking, most companies end up choosing between two options: distribute lots of smaller UPSs in or near their server racks or deploy one to two large UPSs at a central spot in their data center. As anyone who has ever investigated the matter knows, both approaches have their strengths and weaknesses. Distributed deployment schemes tend to deliver better availability and is also easier on capital budgets, as they allow you to add capacity incrementally on a pay-as-you-grow basis. Centralized schemes, on the other hand, are generally easier to manage and don t force you to devote precious rack space to power protection gear instead of servers. So which strategy is right for you? As is so often true in life, there s no easy answer, but the goal of this issue of Data Center Forum is to help make arriving at the correct choice for your particular environment easier. First, in our cover story, Eaton expert, Ed Spears explores the pros and cons of both distributed and centralized power protection architectures and outlines six questions to ask when deciding which one to employ or even whether you should use a combination of both. He also proposes a series of steps any company can take to lessen the downsides associated with whatever architecture they ultimately choose. Next, we offer a case study that examines the distributed versus centralized UPS deployment debate in a real-world context. ICS, a developer and distributor of computer systems for manufacturers, warehouses and construction firms, started out using a distributed power protection scheme before switching to a centralized architecture instead. Why and how they did so makes for interesting reading. Finally, we ve also included a new edition of our IT Interactive section. In addition to the usual array of thought-provoking facts and figures, you ll find a surprising report on the slowing growth of data center power consumption and a quick look at four unusual data centers, including one housed in a tent. I also encourage you to send us your perspective on distributed and centralized power protection approaches for a chance to win a jumbo cooler. Choosing between a distributed and centralized UPS deployment strategy is one of the most basic and important power quality decisions a data center manager must make. I hope the information and insights we share in this issue of Data Center Forum help make arriving at your own answer to that ageold conundrum a little simpler. Register to receive the quarterly electronic version of Data Center Forum: eaton corporation

4 IT Perspective Distributed versus centralized power backup strategies Choosing the right UPS deployment architecture for your data center Though power protection solutions vary in numerous ways, all but the largest data centers use one of two basic approaches when deploying uninterruptible power systems (UPSs): Either they distribute many smaller UPSs among their server racks or they install one or two large UPSs centrally within or near their server room. Both strategies have strengths and weaknesses, and there is no simple formula for determining which one a given company should use. This editorial attempts to simplify the decision-making process, however, by explaining each architecture s pros and cons, outlining the key questions to ask when deciding between them and describing ways to mitigate the weaknesses of whichever approach you ultimately select. Distributed and centralized backup basics Though both are designed to keep important data and applications continuously available, distributed and centralized power backup schemes accomplish that goal in different ways with differing advantages and disadvantages. Distributed backup In a distributed power backup scheme, UPS hardware and other power protection equipment is mounted directly in an organization s server racks, or in some cases directly adjacent to them. Commonly used in network closets and mediumsized data centers, distributed backup strategies generally feature large numbers of smaller UPSs located in many places. Advantages of the distributed backup approach include the following: Lower capital expenses: Distributed backup architectures help growing companies conserve cash by adding UPSs incrementally each time they install a new server rack, instead of making large upfront investments in centralized UPSs with more capacity than they initially need. Furthermore, since the smaller UPSs typically employed in distributed backup solutions tend to be affordably priced, businesses can economically strengthen redundancy by equipping each of their racks with two UPSs rather than one. Finally, the typical lifespan of the compact UPSs employed in distributed backup schemes is roughly the same as the typical lifespan of a server. As a result, organizations with distributed backup architectures can easily amortize and replace their UPS hardware on the same three to five year schedule as their server hardware. Better power availability: As a rule, the shorter the wiring distance between UPSs and the servers they protect, the lower the potential for problems along the power chain (such as grounding issues, electrical noise interference and loose connections) and therefore the higher your overall power availability. Distributed backup strategies locate UPSs in or directly adjacent to server racks. As a result, they provide fewer opportunities for wiring flaws to develop than centralized architectures and generally deliver better availability. Easier deployment and redeployment. Distributed backup architectures generally feature smaller, lighter UPSs that are easy to install and relocate. That can be an advantage for companies that redeploy equipment in their data centers on a frequent basis. Simpler network integration. In distributed backup schemes, UPSs are positioned in or adjacent to server racks, so Ethernet connectivity is always nearby and readily available. Figure 1. Centralized power backup strategies rely on one or two large UPSs 4 eaton corporation

5 IT Perspective Disadvantages of the distributed backup approach include the following: More complex management: Companies with distributed backup strategies often find monitoring the status of their many UPSs a challenge. In addition, administering numerous UPSs in multiple locations can be a demanding and time-consuming chore for data center technicians. Less efficient rack space usage: Though power protection equipment is an essential component of any well-designed IT infrastructure, servers usually play a far more important role. Consequently, most IT managers seek to devote as much rack space as possible to server hardware. In distributed backup strategies, however, UPS hardware takes up rack slots that could otherwise be dedicated to servers. Centralized backup A centralized backup scheme relies on one or two large UPSs rather than lots of smaller ones. Those UPSs are usually situated at the end of a server row, along the perimeter of the server room or in a separate room nearby. Advantages of the centralized backup approach include the following: Easier management: Managing one or two UPSs is far less labor-intensive for administrators than overseeing lots of distributed devices. Simpler redundancy: Adding redundancy to a distributed backup scheme requires companies to install two UPSs in each of their server racks, doubling the already sizeable number of devices they must monitor and manage as well as the amount of rack space they must dedicate to power protection. Redundant centralized power strategies require just two UPSs, rendering them much easier to administer. Better use of rack space: By locating UPSs outside the rack, centralized backup strategies leave more space for server hardware. Three-phase power protection: The smaller UPSs typically used in distributed backup architectures generally provide single-phase power protection only. However, highdensity server equipment such as blade servers often runs on three-phase power. The larger UPSs employed in centralized backup schemes can protect both three-phase and singlephase electrical loads. Longer battery life in hightemperature environments. Thanks to research proving that companies can safely operate servers at temperatures as high as 80ºF/27ºC or above, server rooms are often far warmer these days than was once common. Such hot conditions can significantly shorten the lifespan of a UPS battery. In a centralized backup scheme, batteries are typically located in a cooler space outside the data hall, so they tend to require replacement less frequently. Disadvantages of the centralized backup approach include the following: Higher capital expenses: The large UPSs utilized in centralized backup strategies are generally far more expensive than the smaller devices used in distributed architectures. Lower energy efficiency: To accommodate future growth, most companies with centralized backup schemes deploy UPS hardware with more capacity than they initially require. Consequently, their UPSs are lightly loaded at first and often as little as 75 percent efficient. That means up to 25 percent of the power those UPSs consume ends up being wasted. To make matters worse, all of that wasted electricity is ultimately emitted as heat, putting extra burdens on data center cooling systems and further driving up energy costs. More complicated network integration. In centralized backup schemes, UPSs are positioned outside the data hall in rooms that may not have ready access to Ethernet connectivity. Questions to ask when deciding which backup architecture to use Though there is no easy way to determine whether a distributed or centralized backup strategy makes more sense for your data center, here are six key questions to ask during the decision-making process: 1. Will I be using server hardware that requires three-phase power protection? If so, a centralized power backup scheme is probably your best option, as few of the rack-mounted UPSs used in distributed backup strategies offer three-phase power protection. Weighing the Tradeoffs Deployment Architecture Advantages Disadvantages Distributed Centralized Better power availability Easier redeployment Simpler network integration Lower capital expenses Simpler redundancy Better use of rack space Three-phase power protection Longer battery life in hightemperature environments Easier management Less efficient rack space usage More complex management Lower energy efficiency Higher capital expenses More complicated network integration Six Essential Questions 1. Will I be using server hardware that requires three-phase power protection? 2. Which strategy makes better financial sense? 3. Am I equipped to monitor and manage distributed UPSs? 4. Which approach will better meet my availability needs? 5. How big a priority for my data center is maximizing server density in the white space? 6. Who will be primarily responsible for managing and maintaining our UPSs? eaton corporation

6 IT Perspective 2. Which strategy makes better financial sense? Answering this question involves weighing multiple considerations: Sensitivity to capital expenses: Companies looking to limit capital expenditures tend to prefer distributed backup schemes, since they allow you to buy additional UPS capacity gradually over an extended period of time, rather than all at once upfront. It s worth noting, however, that many growing companies with a distributed backup strategy eventually wind up switching to a centralized architecture when the sheer volume of UPS hardware in their data center becomes unmanageable. In such cases, total capital outlays over multiple years can easily add up to a sum greater than the cost of simply utilizing a centralized approach from the start. Assuming they have enough cash on hand to afford it, organizations with enough confidence in their growth projections to foresee a likely need for large, centralized UPSs within a few years time should carefully assess the wisdom of going straight to a centralized backup approach rather than using a distributed one first. Sensitivity to operating expenses: For businesses eager to limit operating expenditures, centralized backup strategies have both strengths and weaknesses. On the one hand, they usually require less administrative overhead. On the other, their lower initial energy efficiency typically translates into higher electricity spending. Pros and cons with a distributed backup strategy are exactly the opposite: Management costs tend to be higher than with centralized architectures but energy spending tends to be lower. 3. Am I equipped to monitor and manage distributed UPSs? Administration is one of the top challenges associated with distributed backup schemes. Companies considering a distributed approach, then, should evaluate their readiness to monitor and manage potentially large numbers of UPSs scattered across their data center. 4. Which approach will better meet my availability needs? As noted earlier, distributed backup architectures tend to deliver better power availability, since they position UPSs in close proximity to the loads they re protecting and the failure of a single UPS affects only a small portion of the data center. Companies with an extremely low tolerance for downtime such as hospitals and stock trading firms may therefore be better off adopting a distributed backup strategy versus a centralized one. 5. How big a priority for my data center is maximizing server density in the white space? In distributed backup architectures, UPSs and batteries occupy rack space that otherwise could be dedicated to servers. As a result, organizations that wish or need to fit as many servers as possible into their white space usually choose to place UPSs in their grey space by adopting a centralized backup scheme. This is especially true of smaller installations with a limited number of racks. 6. Who will be primarily responsible for managing and maintaining our UPSs? As a rule, facilities managers are more familiar and hence comfortable with centralized backup architectures, while IT managers are more comfortable with distributed ones. For that reason alone, companies that give primary responsibility for UPS management to their facilities department often choose a centralized backup scheme, while companies that place their IT department in charge of UPS administration go with a distributed scheme. Figure 2: State-of-the-art power management software enables data center staff to view and control an entire distributed UPS infrastructure through a single console. Mitigating the downsides of distributed and centralized power backup strategies Though both distributed and centralized backup approaches have flaws, data center operators can lessen their force with the help of these suggestions: Deploy intelligent power management software and network-enabled UPSs Businesses planning to use a distributed backup architecture can make monitoring and managing their numerous UPSs easier by implementing networkaware power management software. When used in conjunction with network- 6 eaton corporation

7 IT Perspective enabled UPSs, the latest intelligent, logical and complete power management applications give data center managers a global view of their entire power quality infrastructure through a single console. Adopt rigorous UPS maintenance procedures Another way to simplify management of distributed power backup architectures is to employ thorough, documented maintenance procedures, such as those contained in the Information Technology Infrastructure Library (ITIL). Developed by the British government in the 1980s, ITIL defines specific, effective and repeatable ways to handle incident management, service desk operation and other common IT tasks. Even in data centers with large numbers of Figure 3: Like a blade server, modular UPSs let data centers add capacity incrementally as their needs increase. Available space for additional UPS power capacity and battery runtime UPSs, performing administrative tasks in consistent, repeatable ways can significantly lower the likelihood of power failures while also increasing the productivity of data center technicians. Use UPSs that are energyefficient at low loads One drawback of centralized backup strategies is that the large UPSs they rely on are often lightly loaded at first and hence inefficient. However, companies can reduce the resulting wasted energy and heightened cooling demands by deploying UPSs designed to be efficient even under low loading conditions. UPSs with state-ofthe-art, multi-mode conversion capabilities, for example, are highly efficient even when less than fully utilized. Utilize modular UPSs Building your power protection architecture around modular UPSs can be a best of both worlds solution to the distributed versus centralized backup dilemma. Much like blade servers, modular UPSs enable multiple plug-and-play power units to share common electrical feeds, power supplies and cabling. Some designs provide up to 50 or 60 kw of capacity in 12 kw building blocks; larger ones can be augmented in 200 to 300 kw increments. Either way, data centers can simply plug in another power module whenever their requirements increase. The end result is a redundant backup scheme endowed with both the manageability of a centralized strategy and the affordable, pay-as-you-grow scalability of a distributed approach. Employ a hybrid backup strategy Distributed backup strategies improve power availability by placing UPSs close to servers. Centralized backup strategies, however, offer a simpler way to add redundancy to a power protection scheme. Hybrid backup strategies that utilize both a large, centralized UPS and numerous smaller, distributed UPSs combine the best of both approaches. Should a problem occur along the power chain between the centralized UPS and the racks it safeguards, the distributed UPSs can provide continued protection. Conversely, should one or more distributed UPSs fail during an electrical outage, the centralized UPS is there to keep emergency power flowing. Conclusion When designing a power protection scheme for their data center, IT and facilities managers must ask themselves whether a distributed or centralized backup strategy makes more sense. Unfortunately, there is no easy answer to that question. Companies must weigh each architecture s advantages and disadvantages against their financial constraints, availability needs and management capabilities before deciding which one to employ. By following the suggestions in this white paper, however, data center operators can not only simplify their decision-making process but lessen the potential weaknesses of whichever strategy they ultimately select. Of course, deploying both centralized and distributed backup devices costs more in the short run than choosing one approach or the other, but organizations with high availability requirements are likely to find the added expense worthwhile in the long run. eaton corporation

8 Customer Voice A Developmentally Sound Power Protection Plan Background For more than 25 years, ICS has been providing advanced software solutions delivering superior business performance to wholesale distributors, manufacturers, public/third-party warehouses and construction firms. The company which operates as four business units that share common management, operations, marketing and knowledge resources to achieve divisional objectives offers two leading brands of enterprise (ERP) software: Microsoft's Dynamics AX and Oracle's JD Edwards. ICS is also the award-winning developer of RF-SMART ADC and RFID solutions for JD Edwards and Dynamics AX, as well as Equential mobile applications for the construction industry and Logimax WMS solutions for public/third-party warehouse management. Challenge Developing software within a highly virtualized environment demands continuous uptime. If the power were to fail at ICS, We would have a lot of expensive developers sitting around doing nothing completely wasted productivity, reveals Chris Bak, the company s IT manager. It s extremely important that our systems are up and available. As a result, when ICS moved in to a new facility in the summer of 2005, the firm sought an uninterruptible power system (UPS) capable of delivering the highest level of reliability and complete protection to its entire data center, which includes more than 50 servers. Having previously relied on numerous smaller UPSs, ICS wanted to upgrade to a centralized solution that would afford ease of management while also simplifying maintenance. At the same time, the company required a UPS that would not occupy a lot of space on its data center floor. Company: ICS Location: Jacksonville, FL Segment: Software development Application: Power protection Challenge: Replace multiple, smaller uninterruptible power systems with a single bigger unit that would simplify management without occupying large amounts of floor space. Solution: An Eaton 9155 UPS and ongoing Eaton service plan Results: Thanks to its new power protection solution and service plan, ICS is now equipped to: Keep mission-critical servers continuously available Conserve valuable data center floor space Perform maintenance procedures without powering down connected loads Receive outstanding service and support from skilled Eaton engineers 8 eaton corporation

9 Customer Voice We wanted to take up the least amount of real estate possible on our data center floor this unit was able to fit in perfectly. Chris Bak, IT Manager Solution Recommended to ICS by one of its equipment representatives, the Eaton 9155 UPS met all of the company s specifications and has been impressing Bak since day one. The system quite literally sits in the corner and we don t touch it, he explains. That, to me, is the most important function. Out of site is out of mind, and it truly just sits there and does its job 24x7. And it s going on five years old now. The unit s online design completely isolates equipment from utility power to protect against all nine of the most common power problems, including outages, sags, surges, spikes, brownouts, line noise, frequency variation, switching transients and harmonic distortion. At the same time, a high 0.9 output power factor delivers more power than the vast majority of competitors models, enabling the 9155 to power more equipment than other UPSs of equivalent VA rating. We wanted a system that could provide complete power to our data center, given our requirements, notes Bak. The 9155 was able to do that with the power it provided. ICS is also benefiting from the 9155 s high efficiency rating of 90 percent across all load ranges, which lowers utility costs and allows the unit to runs cooler, thereby extending component life. Another key factor for ICS was the unit s space-saving, compact and attractive design, which offers high power density per square foot. The sleek tower model measures just 12 inches wide and 33 inches deep. Even more, unlike competitors models, the 9155 s design includes the batteries. It was important that the UPS take up the least amount of real estate possible on our data center floor, Bak explains. This unit was able to fit in perfectly. The centralized solution has significantly eased management and maintenance hassles for ICS. It allows us to monitor the entire room from one source, Bak explains. As a result, I m not running around replacing batteries every three months on a bunch of smaller units. The maintenance is all on one schedule, which is very handy. Further simplifying management for ICS is Eaton s NetWatch Client 5.0 UPS connectivity software, which enables the configuration of up to 255 network clients to monitor the activity of a single UPS. During an extended power failure, NetWatch works in conjunction with a ConnectUPS Web/SNMP network card to perform an automatic, graceful shutdown of UPS-protected devices. In addition, NetWatch's robust configuration options allow scheduled shutdown and restart operations while providing immediate user notifications. In our tests, it has worked really well, confirms Bak. The 9155 also affords long battery life, compliments of Eaton s exclusive ABM technology, which significantly increases battery service life. Unlike conventional tricklecharging, ABM uses an innovative, three-stage charging technique that offers prolonged rest periods between charge phases, as well as temperaturecompensated charging to optimize recharge time. In the future, ICS may also take advantage of the 9155 s Hot Sync technology, which allows up to three equivalent UPS modules to be paralleled for extra capacity or redundancy, ensuring that critical systems are always protected. Parallel configurations can provide up to N+2 redundancy, with no inter-module communication wiring needed. This patented load-sharing control distributes the load equally and keeps the load fully protected even if a paralleled UPS is unavailable. The unit gives me a real peace of mind, Bak says of the It allows me to focus on everything else that needs to be done. Implementation When ICS deployed the 9155 within its data center, the company discovered that ease of installation is yet another advantage to the UPS. Not only does the unit s small footprint support more location options, it also provides fast and simple installation, while lowering deployment costs and preserving space for other equipment needs. Following a smooth deployment, ICS opted to preserve the ongoing health of the 9155 with regular preventive maintenance performed through an Eaton service plan. When you re relying on the UPS for your power needs, you want to make sure you don t have an issue that could have been otherwise prevented and could lead to downtime, Bak explains. It s our peace-of-mind insurance plan. With routine maintenance performed annually, Bak has nothing but praise for Eaton Customer Service Engineers (CSEs) Brian Irish and James Jacobson. The guys who come out and work on it are great, Bak enthuses. They are very professional and represent themselves well. They always do a very good job. Results I m extremely happy with the 9155, Bak emphasizes. I would definitely recommend it to any other company looking for a UPS in this size range. With the 9155 safeguarding its critical server environment, ICS can focus its attention on software development, rather than on power protection. With the 9155 in place, the company is able to: Ensure continuous uptime and high availability for its critical equipment Conserve valuable data center space thanks to the 9155 s small footprint Easily perform maintenance without powering down the connected load Extend runtime as needed with additional EBMs Rely on top-of-the-line service from Eaton CSEs eaton corporation

10 Data Center Interactive Brain Teasers: Rack your IT brain Global forum Data Center Energy Usage: Still Growing, but Less Rapidly With the skyrocketing growth of both cloud and mobile computing imposing intense new demands on data centers, electricity consumption by such facilities must surely be accelerating, right? Wrong, according to research by Stanford University professor Jonathan Koomey. In fact, it appears to be slowing down. Between 2000 and 2005, Koomey writes, energy usage by U.S. data centers doubled. Between 2005 and 2010, however, it rose just 36 percent. That s significantly below a 2007 projection from the Environmental Protection Agency, which predicted that electricity use by U.S. data centers would jump 100 percent from 2005 to 2010 just as it did the previous five years, reaching approximately 100 billion kilowatt hours. Why the surprising dropoff? Koomey cites three factors: reduced demand for IT services stemming from the economic crisis of 2008 and 2009, increased use of virtualization technologies, and the data center industry s heightened focus on efficiency. Though Koomey doesn t state conclusively how much credit each of those variables deserves, observers note that any effects produced by the weak economy are likely to vanish when growth picks up steam again. A summary of Koomey s findings and a link to the complete study can be found on his blog at Fast facts Spending on servers, storage and networking equipment by data centers worldwide will reach $106.4 billion in 2012 and rise to $126.2 billion in 2015, according to Gartner Inc. Also from Gartner: In 2015, the top two percent of data centers will account for an estimated 60 percent of floor space and 71 percent of hardware spending worldwide. Approximately 85 percent of companies with $1 billion or more of revenue or at least 5,000 employees plan to expand their data centers in 2012, according to data center solution provider Digital Realty Trust. Just seven percent of data centers globally achieve percent uptime or better, according to analyst firm Aberdeen Group. Survey Send us your response and be one of 25 early bird winners of optical mouse. What topics would you most like to see us cover in future issues of Data Center Forum? Submit Optical Mouse Did you know? Tired of housing your IT infrastructure in the same old brick-and-mortar structures as everyone else? Consider these unconventional alternatives: Healthcare provider Trinity Health placed a collection of servers and networking equipment in a generator shed outside its Novi, Mich., headquarters for the first nine months of Despite wide swings in temperature and humidity, the company says its test site achieved better uptime than Google and Amazon reported during the same time period. Trinity s experiment was inspired by an earlier one conducted by two Microsoft employees who hosted five servers outdoors in a large, metal-framed tent from November 2007 to June After eight months of cold and rainy Pacific Northwest weather, all the devices were running fine. The tent project, of course, was merely a demonstration, but Microsoft plans to house a new production data center in the architectural equivalent of a tractor shed. Located in Quincy, Wash., the simple, barn-like structure will use pre-manufactured containers to house everything from servers and storage to the facility s electrical, mechanical, and security subsystems. Companies looking for something more durable than a tent or barn in which to host mission-critical systems should investigate DEAC s Grizinkalns co-location data center in Riga, Latvia. The facility s 400 square meters of server rooms are 12 meters underground in a former Soviet Army command bunker. 10 eaton corporation

11 Data Center Interactive Glossary Centralized power protection: a backup strategy that relies on a small number of large, centrally deployed UPSs. Distributed power protection: a backup strategy in which UPS hardware and other power protection equipment is mounted directly in or adjacent to an organization s server racks. Hybrid power protection: a backup strategy that uses both distributed and centralized UPS hardware. Information Technology Infrastructure Library (ITIL): a detailed set of procedures for performing IT management tasks in rigorous, efficient and repeatable ways. Modularity: in UPSs, a design approach that allows organizations to add capacity incrementally as their needs increase. Three-phase power: a power distribution technique that separates electrical current across three separate phase wires, allowing higher amounts of power to be delivered to a single point or load. Contest Participate in our new contest for a chance to win a jumbo collapsible cooler. There may be no one correct answer when it comes to choosing between distributed and centralized power protection strategies, but that doesn t stop data center professionals from voicing strong opinions on the topic. Send us your take, and if it s the most compelling, persuasive argument we receive, you ll take home our prize! Submit Jumbo Cooler Contact Us or visit: Eaton Corporation Electrical Sector 1111 Superior Ave. Cleveland, OH United States Eaton, ABM, epdu and Intelligent Power are trade names, trademarks and/or service marks of Eaton Corporation or its subsidiaries and affiliates. All other trademarks are property of their respective owners Eaton Corporation All Rights Reserved Printed in USA DCF_newsv6.1 March 2012 eaton corporation

12 BEFORE HE UPGRADED OUR SERVER ROOM, HE USED TO GIVE US ALL THE EVIL EYE. I MISS THAT. Backup Power (UPS) 99% energy efficient with intuitive LCD display. Rack Power Distribution Manage power consumption. Power Management Software Optimized for virtualized environments. Say goodbye to weekends in the office. Discover our new power management solutions. It s time to reclaim your nights and enjoy your weekends: Things have changed. Eaton offers power management solutions you can count on from UPS products to airflow management systems, rackbased power management to power management software. Your desk toys will just have to keep missing you. switchon.eaton.com