Download Entire 2010 Issue 2 - Dell

power solutions 2010 IssueSolutions for the virtual era DPS-201002-STD02US$18.95 Can$19.95Your guide to maximizing IT efficiencySolutions for the virtual eraHow advanced data center technologiesdrive sustainable growth2010 Issue 02dell.com/powersolutionsThe smart money’son ITDesigning for hyperscalecomputingStorage deduplicationtechnology advancesDell KACE appliances forWindows 7 migrationEnhancing iSCSI SANself-managementPervasive businessintelligence on Oracle

8 cover storyfeaturesSolutions for the virtual eraThe smartmoney’sonITHow theEfficient Data Centerdrives sustainable growthBy Bryan Jones and Bill GoinsImagine a data center where pooled servers,networking, and storage can be rapidlyredeployed to meet changing needs; wherebusiness application developers can selfprovisiontheir workload deployments; and whereoptimized storage and intelligent infrastructure arethe norm. Dell is helping organizations to unlockdata center efficiency and work toward the goal ofreturning as much as 50 percent of the IT budget todrive business innovation.142024Designing for hyperscale computingBy Steven Croce, Brandon Draeger,and Buck AveyCreated for hyperscale environments such asprivate and public clouds, Web 2.0 and onlinegaming infrastructures, and high-performancecomputing, Dell PowerEdge C-Seriesservers provide outstanding performance ina dense, streamlined, eco-efficient design.Advancing server scalabilityand performanceBy Armando Acosta and Robert BradfieldFour-socket Dell PowerEdge R810 andPowerEdge R910 rack servers andPowerEdge M910 blade servers offerhighly scalable processing power andmemory capacity to help cost-effectivelyboost performance and expand virtualizedenvironments, while other features helpincrease reliability and simplify administration.High performance in acompact designMaximizing performance while makingefficient use of available power and space canbe critical in data center environments. TheDell PowerEdge R815 rack server with theAMD Opteron 6000 Series platform deliversfour-processor performance in a compact,cost-effective 2U design.6Fresh outlook,bold new design28How storagededuplicationtechnology pays off56Monitoring Dellstorage in MicrosoftSystem Center96 Near-instantaccess without afull system boot104Boosting businessefficiency runningOracle BI on DellReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved. dell.com/powersolutions | 2010 Issue 02 1

42Tsumura & Company:Solving IT problems withserver consolidationBy consolidating its infrastructure on virtualized DellPowerEdge servers, Japanese herbal medicine leaderTsumura & Company has laid the groundwork for a dramaticreduction in ongoing server costs while ensuring that legacyapplications can continue running years into the future.featuresSolutions for the virtual era2834How deduplication helps reduce thecost of backup and disaster recoveryBy Kay Benaroch and Shane JacksonDell and EMC have teamed up to deliver efficientdisk-based backup and recovery with integrateddeduplication technology designed to make disk-basedbackups as cost-effective as tape-based backups.Optimizing management efficiencywith Dell/EMC NS storageBy Annette Cormier, Eric Cannell, and Brad BunceThe Dell/EMC NS family of unified storage enablesorganizations to easily and cost-effectively consolidateboth unstructured file-system data and structuredapplication data into a single shared storage system—helping organizations control costs, streamlinemanagement, and enhance storage availabilityand performance.44State of Delaware:Taming the e-mailbeast with bladesCentralizing its e-mail systems onDell PowerEdge blade servers andDell EqualLogic storage helps thestate of Delaware reduce total cost ofownership, streamline management, andcut downtime from days to minutes.39Racking up data center efficiencyBy Joyce RuffDesigned to meet the needs of high-density data centerenvironments, the Dell PowerEdge 4820 rack enclosurecan hold and protect a wide range of IT equipmentwhile supporting enhanced power management,efficient cooling, and simplified component storageand mounting.2 2010 Issue 02 | dell.com/powersolutions Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

50Storage optimizationEnhancing SAN self-managementwith Dell EqualLogic storageBy Tony Ansley and Mansour KaramDell EqualLogic arrays offer built-in intelligence to help simplifyconfiguration—but operators may still miss important settings.Together, Arista and Dell have created a storage architecturethat can automate configuration through switches running theArista ® Extensible Operating System (EOS ® ) platform.Business intelligence104 Enabling pervasive businessintelligence using an integratedOracle platformBy Mark Conway, Dan Blankenship,and Jaime DelgadoOEM solutions110 Enabling eco-friendlydeployments with NEI platformsBy Richard Graber114 Custom fit: The Dell PowerEdge R310appliance server platformBy Franklin FlintdepartmentsEditor’s comments6 Designs on efficiencyBy Tom KolnowskiStorage optimization56 Simplified monitoringof Dell EqualLogic storagewith Microsoft System CenterBy Viswanathan Balakrishnan,Stanley L. Stevens, Sekhar Duggirala,and Tom George61 Boost storage consolidationwith Blue Coat WAN Optimizationand Dell PowerVault appliancesBy George Sadler and Suresh Jasrasaria66 F5 ARX file virtualizationand Dell storageBy Renny Shen70 Accelerate and secure datareplication with F5 WAN optimizationBy Fred Johnson, Ujjwal Rajbhandari,and Puneet Dhawan75 A unified networking approach to iSCSIstorage with Broadcom controllersBy Dhiraj Sehgal, Abhijit Aswath,and Srinivas Thodati80 Addressing data growthchallenges with Symantecdeduplication softwareBy Charles Butler84 End-to-end, snapshot-awaredata protection withCommVault SnapProtectBy Darin CampVirtualization88 Overcoming data protectionchallenges in virtualizedenvironmentsBy Scott Herold91 Simplifying networks in virtualizedenvironments with Intel technologyBy Srinivas Thodati and Brian JohnsonMobility solutions96 Productivity at the touch of abutton: Dell Latitude ON | FLASHBy Roberto Ayala and Steve Rokov100 Securing the mobile workforcewith SonicWALL appliancesBy Patrick Sweeneyand Matthew Dieckman46Systems managementDell KACE appliances:SimplifyingWindows 7migrationsBy Sendhil JayachandranFor midsize organizationsworking with a diverse rangeof hardware and limitedresources, making the moveto the Microsoft ® Windows ® 7OS can be a dauntingchallenge. The innovativeDell KACE family ofappliances provides asimplified, automated,cost-effective way to designand carry out an end-to-endmigration.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 3

For more information, see “Designing for Hyperscale Computing,” by Steven Croce, BrandonDraeger, and Buck Avey, in Dell Power Solutions, 2010 Issue 2, dell.com/downloads/global/power/ps2q10-20100360-cloud.pdf.IntelligentinfrastructureSimplifiedinfrastructuremanagementVirtualization-ready Tools for managinginfrastructure for and deployingrapid deployment physical, virtual,and optimized and logicaloperationsinfrastructure• Optimized server • Policy-basedplatformsupdates and• Efficient I/O fabric configuration• Management ofservers, networks,and storage in asingle pool• Physical serversVMwareCitrix ® XenServervirtualizationplatformsRapid deliveryof strategic ITservices withgroundbreakingcapabilities• Drag-and-dropimage creation• Self-serviceworkloaddeployment• AdvancedmonitoringIntelligentData ManagementAutomatedmanagement ofdata throughpredictable,scalable, and opendata storage• Efficient objectstorage• Deduplication• Unified and tieredstoragePower Solutions onlineBrowse our Web site:dell.com/powersolutionsGet the Digital Edition:powersolutionsdigital.dell.comFollow us on Twitter:twitter.com/powersolutionsonline exclusiveStreamlining systems managementwith Dell Management Console 1.1By Todd Rodgers and Tad WalshDell Management Console powered by Altiris from Symantec consolidates multiple systemsmanagement tasks into a unified, easy-to-useconsole. Dell Management Console 1.1 enhancesthis functionality with features including powermonitoring, out-of-band update capabilitiesfor server BIOS and firmware, and automatedwarranty reporting.Dell Power Solutions Magazine and specialedition articles are also available online atdell.com/powersolutions. Check theDell Power Solutions Web site for early-releasearticles, how-to’s, case studies, best practices,and expert tips you won’t find anywhere else.Power Solutions Digital EditionDell Power Solutions Digital Edition offers a comprehensive IT library at yourfingertips—featuring powerful searches across archived issues and direct links torelated resources. For instant accessibility, visit powersolutionsdigital.dell.com.FeaturesectionSolutions for the virtual era“Dell Advanced InfrastructureManager software allowsIT departments to managenetworking, storage, and servers—both physical and virtual—as a single resource pool.”Best practices: Envisioning theEfficient Data CenterIn the Dell Efficient Enterprise vision, technologyinfrastructures are open, capable, and affordable.Products and services for the data center are flexibleand avoid locking organizations into using a singlevendor or proprietary technologies that contributeto increased TCO and limit future choice; productsand services have innovative functionality that meetsexisting requirements and provides the flexibility toadapt to future requirements. In addition, Dell’s visionblends new technology with existing infrastructure,existing expertise, and industry-standard approachesdesigned to drive down TCO and complexity.Dell begins with a three-step approach to helporganizations unlock enterprise efficiency andachieve significant returns on their IT investment.The first step on the journey is to optimize theexisting data center infrastructure, with the goal ofachieving uniformity and cutting costs. Commonplatforms, standards-based tools, and unifiedfabrics can contribute to these efforts. The goal: toeliminate proprietary legacy architectures, like RISCandSPARC-based UNIX ® systems, and replace themwith a scaled x86 architecture running Microsoft ®Windows ® or Linux® operating systems to help lowercosts and reduce management oversight.The next step is to simplify the technologyinfrastructure. Using pragmatic approachessuch as virtualization and storage consolidation,organizations can consolidate or unify redundanciesto help ensure that they are getting the most outof their data center resources. Rationalizing andreducing the number of applications also enhancesthe flexibility to easily remove and repurpose servers.After the data center has been standardizedand simplified, enhanced levels of automationbecome possible. The third step is to implementbest practices that help reduce manual interventionand boost productivity, together with cloud-baseddelivery models (where appropriate) that help reduceadministrative burdens on IT staff and enhanceapplication availability. Managed service options alsoallow organizations to combine 24/7 monitoring,alerts, and reporting with expert analysis and advice.In the Dell model, the Efficient Data Center isbuilt on four key pillars (see Figure 1).Intelligent infrastructureDell’s Efficient Data Center approach can helporganizations optimize their existing data centers,virtualize on their own schedule, take advantage ofcloud technologies as they make business sense—and prepare their data centers for the future.Achieving these goals requires advanced servers,networking equipment, and storage. Intelligentinfrastructure is designed to automate frequenttasks to drive down TCO, rapidly respond to changerequests, and help administrators anticipate issuesand proactively respond. Dell offers a range ofinfrastructure products to help meet these needs.Purpose-built hardware, like Dell PowerEdgeC-Series servers optimized for cloud applications,can help administrators accomplish specific tasks.An efficient fabric helps lower costs and allowsadministrators to rack and cable only once. Cloudservices enable applications to be cloud optimized.Intelligent infrastructure helps optimize the datacenter today and build the foundation for advancedEfficient Data Center services.10GbE data center fabricIn this Webinar, learn how innovative technologies such asvirtualization and 10GbE enable organizations to converge storagetraffic onto a common unified data center fabric. The recordedpresentation explores how integrating IT management, server,storage, and networking solutions can help accelerate 10GbEdeployments—enhancing flexibility and performance whilereducing cost and complexity.eseminarslive.com/c/a/IT-Infrastructure/Dell041610Dell gears its approach toenterprise efficiency towardtaking advantage of thehardware and software alreadyin the data center while alsoincorporating purpose-builthardware. With the right mixof blade servers, rack servers,towers, and custom formfactors, IT departments canachieve optimum levels ofcompute density, memory,and I/O to help meet specificenterprise needs.Dell PowerEdge servers aredesigned to support generalizedworkloads, hardware-basedavailability, and a traditionalbreak/fix service model.PowerEdge C-Series servers arestandardized and purpose-builtfor cloud applications. 1 And DellData Center Solutions (DCS)are designed to support thelargest public cloud providersin the industry with specializedworkloads, application-basedavailability, and custom servicemodels.10 Gigabit Ethernet (10GbE)connectivity is rapidly becomingthe industry standard fornetworking in the data center.Enterprises can leverage thisconvergence to help reduceoperating costs and avoidcomplexity while continuingto extract value from existinginvestments. Because the 10GbEstandard allows organizationsto continue using their existingplatforms and networkingtools, they can extend capitalinvestments and IT staff trainingto the updated infrastructure.10GbE bandwidth also allows thereduction of port count to helpreduce cost, while a multi-vendorapproach helps preserve choicein adopting new technologies asthey become available.Key Dell connectivityofferings include DellPowerConnect switches,which provide Gigabit Ethernet(GbE) and 10GbE rack andblade switches that can deliverextreme price/performancetargeted for edge networkapplications. Dell PowerConnectB-Series switches offer GbE and10GbE rack and chassis switchesfor price-sensitive aggregation,core, and storage area networkFour pillars:Building theEfficient Data CenterIntelligent infrastructure: Deploys servers, networking,and fabrics that are designed and purpose-built to helpreduce acquisition and operating costs, with the goals ofreducing the number and types of devices to manage, rackingand cabling once, taking advantage of the 10GbE convergence,and maintaining the flexibility of a multi-vendor approachSimplified infrastructure management: Leveragestechnologies that enable IT staff to consolidate toolsand pool resources, helping them to increase the operatingefficiency of both new and existing infrastructures and todynamically redeploy assets as conditions changeStreamlined application and workload management:Includes tools that allow rapid delivery of strategic ITservices with the goals of enabling application developersto easily create images and self-provision their workloaddeployments, automating approval and deployment processes,and providing self-service cataloged services and chargebacksDell Intelligent Data Management: Optimizes contentstorage throughout its life cycle by automatically placingit on the appropriate tier and storage type while optimizingstorage for virtualized environmentsand MicrosoftHyper-V ,®, andStreamlinedapplicationand workloadmanagementFigure 1. Four key pillars support the Dell model for building the EfficientData Center10 2010 Issue 02 | dell.com/powersolutions dell.com/powersolutions | 2010 Issue 02 11Efficient EnterpriseframeworkTake control of your data center with open, capable,and affordable solutions that help you focus on drivingbusiness innovation rather than just “keeping the lightson.” For the full story, visit efficiententerprise.com.Join the Dell TechCenter community atdelltechcenter.com to share your experiences,voice your concerns, and pick the brain of your favoriteDell expert—or be the expert by signing up for theDell Masters program. And you won’t want to missTechTuesday chats, community forums, discussion groups,blogs, white papers, video demos, and much more.Index to featured Dell customersNAB Solutions............................. 17R Systems..................................... 15State of Delaware...................... 44Tsumura & Company................42Index to advertisersAdvanced Micro Devices, Inc... 7American Power ConversionCorporation............................33Broadcom Corporation.............. 5Dell Inc. ................................99, C3Eaton Corporation.....................55Extreme Networks, Inc.............45F5 Networks, Inc........................27Intel Corporation ....................... 41Microsoft Corporation.............C2NetScout Systems, Inc.............32Network Engines, Inc.............. 113Oracle Corporation..................C4SonicWALL, Inc...........................53Vizioncore Inc.............................654 2010 Issue 02 | dell.com/powersolutions Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Editor’s commentsDesigns onefficiencyWhat began life as, quite literally, a bulleted list of core publishing designtenets and the clean slate of a digital sketch pad has metamorphosedinto the redesigned IT magazine you are holding in your hands, viewingonline, or scanning from the frames of our Digital Edition.Once we began designing, no pixel in our page layouts was left unscathed. The resultis an end-to-end transformation of Dell Power Solutions, into what we hope you will findto be a bold new visual design that is at the same time information-rich yet more openand accessible—so you can more efficiently zero in on the information you need bothwithin and outside the borders of the magazine.Beyond the contemporary, Dell-standard Museo for Dell and Museo Sans for Delltypography throughout; fresh color palette; and rethought artwork accents layeredupon the framework of a fresh editorial design system, we’ve also redesigned thelogo on the front cover and included a tagline to reflect our core editorial mission:“Your guide to maximizing IT efficiency.”“Once we began designing, no pixel inour page layouts was left unscathed.”In addition, the margins of our pages have been architected to be eminentlyscannable for links to online content and communities: look for the gray iconographicsignposts to instantly engage with related technology demos, videos, blogs, chats,peer-to-peer discussions, and more. Of course, we think this is most efficientlyaccomplished from our Digital Edition at powersolutionsdigital.dell.com, accessiblefrom any Web browser and from many mobile devices.While our look has decidedly changed, the team behind the scenes has not: specialthanks to managing editor Debra McDonald and art director David Chan for their tirelessdedication to editorial and creative excellence while keeping us on a frenzied productionschedule throughout the redesign process, aided by a gold-class supporting cast includingKathryn White, Jim Duncan, Dawn Davidson, Terrence O’Donnell, James Hurd, Amy J. Parker,Cissy Walker, Lalaine Gagni, Margie Preston, and Cynthia Webb.Editorial staffEditor-in-chief and publisher Tom KolnowskiManaging editor Debra McDonaldFeatures editor Kathryn WhiteAssociate managing editor Jim DuncanSenior editors Dawn Davidson, James Hurd, andTerrence O’DonnellEditorial assistant Amy J. ParkerArt director and cover designer David ChanDesigners Jonathan Evans, Lalaine Gagni, Matthew Golden,Margie Preston, and Cynthia WebbBusiness development manager Cissy WalkerStaff writers Romy Bauer, Jeanne Feldkamp, Julie Jervis,Greg Thomas, and Chris YoungContributing writers Armando Acosta, Tony Ansley,Abhijit Aswath, Buck Avey, Roberto Ayala, ViswanathanBalakrishnan, Kay Benaroch, Dan Blankenship, RobertBradfield, Brad Bunce, Charles Butler, Darin Camp, EricCannell, Mark Conway, Annette Cormier, Steven Croce,Jaime Delgado, Puneet Dhawan, Matthew Dieckman,Brandon Draeger, Sekhar Duggirala, Franklin Flint, TomGeorge, Bill Goins, Richard Graber, Scott Herold, ShaneJackson, Suresh Jasrasaria, Sendhil Jayachandran, BrianJohnson, Fred Johnson, Bryan Jones, Mansour Karam,Ujjwal Rajbhandari, Todd Rodgers, Steve Rokov, JoyceRuff, George Sadler, Dhiraj Sehgal, Renny Shen, Stanley L.Stevens, Patrick Sweeney, Srinivas Thodati, and Tad WalshContributing photographers Tony Bolding, Bryan Kuntz,Adran Matte, Joey Pena, and Bryce VickmarkAdvertising salesSales director Kyle Walkenhorst (323-340-8585)National sales manager Shaun Mehr (949-923-1660)Western U.S. and South/Central America sales Melany Galley(949-481-1125)Eastern U.S. sales Steve Branda (201-483-7871)Canada, EMEA, and APJ sales Mark Makinney (805-709-4745)Advertising assistant Scott Hallquist (323-254-0905)Ad coordinator Kathy HartloveReader servicesSubscriptions are complimentary to qualified readerswho complete the online subscription form. For the PrintEdition: To sign up as a new subscriber or change yourexisting subscription, access the Subscription Center atdell.com/powersolutions. For the Digital Edition:To sign up as a new subscriber or convert yourPrint Edition subscription to the Digital Edition, visitpowersolutionsdigital.dell.com. For other subscriptionservices, please e-mail us_power_solutions@dell.com.About DellDell Inc., headquartered in Round Rock, Texas, near Austin,listens to its customers and delivers innovative technologyand services they trust and value. Uniquely enabled by itsdirect business model, Dell is a leading global systems andservices company and No. 34 on the Fortune 500 list. Formore information, visit our Web site at dell.com.Dell cannot be responsible for errors in typography or photography.Dell, the Dell logo, Dell KACE, Dell OpenManage, DellPrecision, EqualLogic, Latitude, Latitude ON, PowerConnect,PowerEdge, PowerVault, and ReadyRails are trademarks of DellInc. Other trademarks and trade names may be used in thispublication to refer to either the entities claiming the marksand names or their products. Dell disclaims any proprietaryinterest in the marks and names of others.Dell Power Solutions is published quarterly by Dell Inc.,Dell Power Solutions, One Dell Way, Mail Stop RR5-05, RoundRock, TX 78682, U.S.A. No part of this publication may bereprinted or otherwise reproduced without permission fromthe editor-in-chief. Dell does not provide any warranty as tothe accuracy of any information provided through Dell PowerSolutions. Opinions expressed in this magazine may not bethose of Dell. The information in this publication is subject tochange without notice. Any reliance by the end user on theinformation contained herein is at the end user’s risk. Dell willnot be liable for information in any way, including but notlimited to its accuracy or completeness. Dell does not acceptresponsibility for the advertising content of the magazine orfor any claims, actions, or losses arising therefrom. Goods,services, and/or advertisements within this publicationother than those of Dell are not endorsed by or in anyway connected with Dell Inc.Copyright © 2010 Dell Inc. All rights reserved. Printed in the U.S.A.Printed on recycled paper containing 10 percentpost-consumer waste. Please recycle this magazine.2010 Issue 26 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

4P Performanceat 2P Economics1High-performance features, optimal increased density,and great performance-per-dollar.The AMD Opteron 6000 Series platform helps Dell PowerEdge R815 serverspack a whole lot of punch into a slim 2U form factor, with unprecedented price per processor core. 2The AMD Opteron 6000 Series platform delivers as many as 48 total cores in up to a 4-processor configuration atsignificantly lower prices per core. 2 Featuring the industry’s first 12-core x86 processor, these processors enableup to twice the performance of our previous generation 3 and up to double the cores at the same price. 4Dell PowerEdge R815 servers deliver real world performance for virtualization, databases, and HPC—with up to48 cores and high memory capacities. They’re designed on a consistent architecture that enables scalabilityfor today and tomorrow, plus energy efficiency that can help you conserve precious resources and reduceoperating costs. More than a server, it’s an investment in your future.Learn more at www.dell.com/poweredge/amd.1 Based on standard power Six-Core AMD Opteron processor model 2435 1ku pricing of $989 as of 10/19/09 vs. standard power AMD Opteron processor model 6172 (Magny-Cours) 1ku planned pricing of $989 as of 3/29/2010.2 Based on 1ku pricing for AMD Opteron processor Model 6128 at 3/29/2010 ($266 ÷ 8 cores = $33/core).3 Internal testing at AMD performance labs as of 3/17/10 showed a 2.19x performance gain for 2x AMD Opteron processor Model 6174 (“Magny-Cours”) over 2x Six-Core AMD Opteron processor Model 2435 (“Istanbul”) running SPECfp®-rate2006. Configuration: 2 x Six-Core AMDOpteron processors Model 2435 in Supermicro A+ Server 1021M-UR+B server, 32GB (8x4GB DDR2-800) memory, SuSE Linux® Enterprise Server 10 SP1 64-bit and 2 x AMD Opteron processors Model 6174 in HP ProLiant DL165 G7 server, 64GB (16 x 4GB DDR3-1333) memory, RedHat Enterprise Linux® Server release 5 Update 4 64-bit.4 Based on comparison of Six-Core AMD Opteron processor Model 2435 (6 cores; $989 suggested 1ku price) vs. AMD Opteron processor Model 6172 (12 cores; $989 1ku price as of 3/39/2010).© 2010 Advanced Micro Devices, Inc. All rights reserved. AMD, the AMD Arrow logo, AMD Opteron, and combinations thereof are trademarks of Advanced Micro Devices, Inc. Other names are for informational purposes only and may be trademarks for their respective owners.

FeaturesectionSolutions for the virtual eraThesmartmoney’sonITHow theEfficient Data Centerdrives sustainablegrowthBy Bryan Jones and Bill Goins8 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Imagine a data center where pooled servers, networking,and storage can be rapidly redeployed to meet changingneeds; where business application developers canself-provision their workload deployments; and whereoptimized storage and intelligent infrastructure arethe norm. Dell is helping organizations to unlock datacenter efficiency and work toward the goal of returningas much as 50 percent of the IT budget to drivebusiness innovation.Today, efficiency is a basic tenet of business survival. The boardroom IT conversationis no longer restricted to cutting costs—it’s about investing in technology toefficiently drive business and organizational success.Although IT supports virtually every business process, and tough economic timeshave led to intensified competition, many organizations are hard-pressed to take advantageof technology advances when they are trapped into spending most of a spartan IT budget onmaintenance instead of innovation. As a result, nonstrategic tasks are occupying an increasingproportion of the workday in many data centers.Virtualization and consolidation can go a long way toward driving data center efficiency,but they are just the beginning. Technologies for flexible data management, self-serviceworkload creation and deployment, resource pooling, and smart, self-aware infrastructurehelp to simplify data center management and significantly lower total cost of ownership(TCO). In an environment where IT leaders are looking to partner with business units in away that increases automation and productivity across the organization, advanced data centertechnologies can make these goals a reality.Most technology vendors are focused on tackling the simplest of these challenges byvirtualizing, consolidating, and automating the data center. These vendors typically proposea collection of proprietary offerings designed to lock organizations into vertically integrated,premium-priced infrastructures.Dell offers a model for tackling these challenges and building a solid foundation forenterprise efficiency that does not lock organizations into proprietary solutions—a model thatcan optimize the existing data center, maintain choice in vendors and technology, and allowincremental implementation as time and budget allow. The Efficient Data Center is a keycomponent of Dell’s Efficient Enterprise model, purpose-built to lower costs and increaseIT efficiency.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved. dell.com/powersolutions | 2010 Issue 02 9

FeaturesectionSolutions for the virtual era“Dell Advanced InfrastructureManager software allowsIT departments to managenetworking, storage, and servers—both physical and virtual—as a single resource pool.”Best practices: Envisioning theEfficient Data CenterIn the Dell Efficient Enterprise vision, technologyinfrastructures are open, capable, and affordable.Products and services for the data center are flexibleand avoid locking organizations into using a singlevendor or proprietary technologies that contributeto increased TCO and limit future choice; productsand services have innovative functionality that meetsexisting requirements and provides the flexibility toadapt to future requirements. In addition, Dell’s visionblends new technology with existing infrastructure,existing expertise, and industry-standard approachesdesigned to drive down TCO and complexity.Dell begins with a three-step approach to helporganizations unlock enterprise efficiency andachieve significant returns on their IT investment.The first step on the journey is to optimize theexisting data center infrastructure, with the goal ofachieving uniformity and cutting costs. Commonplatforms, standards-based tools, and unifiedfabrics can contribute to these efforts. The goal: toeliminate proprietary legacy architectures, like RISCandSPARC-based UNIX ® systems, and replace themwith a scaled x86 architecture running Microsoft ®Windows ® or Linux ® operating systems to help lowercosts and reduce management oversight.The next step is to simplify the technologyinfrastructure. Using pragmatic approachessuch as virtualization and storage consolidation,organizations can consolidate or unify redundanciesto help ensure that they are getting the most outof their data center resources. Rationalizing andreducing the number of applications also enhancesthe flexibility to easily remove and repurpose servers.After the data center has been standardizedand simplified, enhanced levels of automationbecome possible. The third step is to implementbest practices that help reduce manual interventionand boost productivity, together with cloud-baseddelivery models (where appropriate) that help reduceadministrative burdens on IT staff and enhanceapplication availability. Managed service options alsoallow organizations to combine 24/7 monitoring,alerts, and reporting with expert analysis and advice.In the Dell model, the Efficient Data Center isbuilt on four key pillars (see Figure 1).Intelligent infrastructureDell’s Efficient Data Center approach can helporganizations optimize their existing data centers,virtualize on their own schedule, take advantage ofcloud technologies as they make business sense—and prepare their data centers for the future.Achieving these goals requires advanced servers,networking equipment, and storage. Intelligentinfrastructure is designed to automate frequenttasks to drive down TCO, rapidly respond to changerequests, and help administrators anticipate issuesand proactively respond. Dell offers a range ofinfrastructure products to help meet these needs.Purpose-built hardware, like Dell PowerEdge C-Series servers optimized for cloud applications,can help administrators accomplish specific tasks.An efficient fabric helps lower costs and allowsadministrators to rack and cable only once. Cloudservices enable applications to be cloud optimized.Intelligent infrastructure helps optimize the datacenter today and build the foundation for advancedEfficient Data Center services.10 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

10GbE data center fabricIn this Webinar, learn how innovative technologies such asvirtualization and 10GbE enable organizations to converge storagetraffic onto a common unified data center fabric. The recordedpresentation explores how integrating IT management, server,storage, and networking solutions can help accelerate 10GbEdeployments—enhancing flexibility and performance whilereducing cost and complexity.eseminarslive.com/c/a/IT-Infrastructure/Dell041610Four pillars:Building theEfficient Data CenterIntelligent infrastructure: Deploys servers, networking,and fabrics that are designed and purpose-built to helpreduce acquisition and operating costs, with the goals ofreducing the number and types of devices to manage, rackingand cabling once, taking advantage of the 10GbE convergence,and maintaining the flexibility of a multi-vendor approachDell gears its approach toenterprise efficiency towardtaking advantage of thehardware and software alreadyin the data center while alsoincorporating purpose-builthardware. With the right mixof blade servers, rack servers,towers, and custom formfactors, IT departments canachieve optimum levels ofcompute density, memory,and I/O to help meet specificenterprise needs.Dell PowerEdge servers aredesigned to support generalizedworkloads, hardware-basedavailability, and a traditionalbreak/fix service model.PowerEdge C-Series servers arestandardized and purpose-builtfor cloud applications. 1 And DellData Center Solutions (DCS)are designed to support thelargest public cloud providersin the industry with specializedworkloads, application-basedavailability, and custom servicemodels.10 Gigabit Ethernet (10GbE)connectivity is rapidly becomingthe industry standard fornetworking in the data center.Enterprises can leverage thisconvergence to help reduceoperating costs and avoidcomplexity while continuingto extract value from existinginvestments. Because the 10GbEstandard allows organizationsto continue using their existingplatforms and networkingtools, they can extend capitalinvestments and IT staff trainingto the updated infrastructure.10GbE bandwidth also allows thereduction of port count to helpreduce cost, while a multi-vendorapproach helps preserve choicein adopting new technologies asthey become available.Key Dell connectivityofferings include DellPowerConnect switches,which provide Gigabit Ethernet(GbE) and 10GbE rack andblade switches that can deliverextreme price/performancetargeted for edge networkapplications. Dell PowerConnectB-Series switches offer GbE and10GbE rack and chassis switchesfor price-sensitive aggregation,core, and storage area networkSimplified infrastructure management: Leveragestechnologies that enable IT staff to consolidate toolsand pool resources, helping them to increase the operatingefficiency of both new and existing infrastructures and todynamically redeploy assets as conditions changeStreamlined application and workload management:Includes tools that allow rapid delivery of strategic ITservices with the goals of enabling application developersto easily create images and self-provision their workloaddeployments, automating approval and deployment processes,and providing self-service cataloged services and chargebacksDell Intelligent Data Management: Optimizes contentstorage throughout its life cycle by automatically placingit on the appropriate tier and storage type while optimizingstorage for virtualized environmentsIntelligentinfrastructureVirtualization-readyinfrastructure forrapid deploymentand optimizedoperations• Optimized serverplatforms• Efficient I/O fabricSimplifiedinfrastructuremanagementTools for managingand deployingphysical, virtual,and logicalinfrastructure• Policy-basedupdates andconfiguration• Management ofservers, networks,and storage in asingle pool• Physical serversand MicrosoftHyper-V ,VMware ® , andCitrix ® XenServer virtualizationplatformsStreamlinedapplicationand workloadmanagementRapid deliveryof strategic ITservices withgroundbreakingcapabilities• Drag-and-dropimage creation• Self-serviceworkloaddeployment• AdvancedmonitoringIntelligentData ManagementAutomatedmanagement ofdata throughpredictable,scalable, and opendata storage• Efficient objectstorage• Deduplication• Unified and tieredstorage1 For more information, see “Designing for Hyperscale Computing,” by Steven Croce, BrandonDraeger, and Buck Avey, in Dell Power Solutions, 2010 Issue 2, dell.com/downloads/global/power/ps2q10-20100360-cloud.pdf.Figure 1. Four key pillars support the Dell model for building the EfficientData CenterReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 11

FeaturesectionSolutions for the virtual era(SAN) applications, while otherupcoming Dell PowerConnectswitches are expected tofacilitate performance-orientedswitching for wide areanetworking and networkingsecurity applications.The Dell Lifecycle Controller,delivered as part of theIntegrated Dell RemoteAccess Controller (iDRAC)Express in 11th-generation DellPowerEdge servers, embedssystems management featuresdirectly in the server—helpingavoid media-based delivery ofsystems management toolsand utilities. This delivery modelhelps simplify provisioning,deployment, patching andupdates, servicing, and usercustomization. It also helps toreduce the time required toaccomplish common tasks,reduce the potential for error,enhance security, and contributeto increasingly efficientmanagement.Simplified infrastructuremanagementManaging servers, networks,and storage as a common poolenables these resources tobe rapidly redeployed to helpmeet dynamically changingneeds. Dell’s Efficient DataCenter approach consolidates,streamlines, and automates themanagement of data centerresources to bring virtual-likefunctionality to physical servers,help eliminate unnecessarymanagement consoles, anddynamically reprovision servers.Advanced Dell systems canhelp administrators to efficientlyoperate heterogeneous datacenters—enabling them torack and cable once, and thenredeploy when needed.Instead of replacingheterogeneous technologyinfrastructures with singlevendorplatforms, Dell focuseson providing comprehensive,streamlined support for therange of platforms already atwork in enterprise data centers(see Figure 2). This approach—which accommodates dynamicmanagement of mixedtechnology environments,including platforms fromMicrosoft, VMware, Citrix, IBM,Hewlett-Packard, and Sun, inaddition to Dell hardware—helpsto deliver dynamic workloadflexibility, enable automateddeployment, and supporthighly efficient operationsmanagement.Dell offers a range oftechnologies designed tosimplify systems management.Dell Advanced InfrastructureFigure 2. A comprehensiveframework of purpose-builthardware and managementtools based on open standardsavoids vendor lock-inManagementframework(Dell AdvancedInfrastructureManager and DellManagementConsole)Manager software allowsIT departments to managenetworking, storage, andservers—both physical andvirtual—as a single resourcepool. This approach helpsavoid resource silos, allowingorganizations to allocateresources dynamically to meetchanging workload demands.In addition, IT managers canseamlessly migrate workloadsbetween physical andvirtual environments to helporganizations meet virtualizationgoals and extend the use ofvirtualization technology in theenterprise while also avoidingduplicate tools and processes.Dell Management Consolepowered by Altiris fromSymantec provides a singlecomprehensive view into anorganization’s IT infrastructureto help reduce the managementcomplexity traditionallyassociated with data centerinfrastructure. IT managers canuse this tool to monitor theServervirtualizationServersI/OvirtualizationNetworkingStorageMicrosoftHyper-VDell PowerEdgeM-Series blade serversDell NX4 networkattached storage(NAS) systemshealth of key systems or systemsubcomponents, and performbasic provisioning, discovery,inventory, power control, andalerting tasks.Streamlined applicationand workload managementAdvanced, application-awareworkload and service monitoringas well as on-demandmonitoring of underlyingphysical and virtual platformscan help dramatically reduceadministrative burdens on ITstaff—freeing them to focus onstrategic projects that add valueby advancing business goalsand organizational objectives.In addition, by automatingand standardizing the wayIT resources are deployed,organizations can reduce thetime and manual interventionrequired to allocate resources toapplications. Tiered infrastructureofferings further help to createuniformity and facilitate costcontainment and chargebacks.VMwareDell PowerEdgerack serversXsigo I/O Director devicesDell PowerConnect switchesDell EqualLogicPS Series iSCSI SANsCitrixXenServerDell PowerEdgeC-Series serversDell/EMCFibre Channel SANs12 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

FeaturesectionSolutions for the virtual eraDesigning forhyperscale computingBy Steven Croce, Brandon Draeger, and Buck AveyCreated for hyperscale environments such as private andpublic clouds, Web 2.0 and online gaming infrastructures,and high-performance computing, Dell PowerEdge C-Series servers provide outstanding performance in adense, streamlined, eco-efficient design.For many organizations buildinghyperscale IT infrastructures, theirdata centers are their factories inWeb 2.0, online gaming, searchengine, and high-performance computing(HPC) environments. Running hundreds orthousands of servers, these organizationsneed to maximize data center efficiency atevery turn. With each infrastructure decision,these organizations are keenly aware of theimplications for their bottom line.To help maximize efficiency and reducethe cost of these hyperscale environments,organizations often require something differentthan traditional general-purpose servers. Theyneed streamlined systems without the extensivemanagement tools, enterprise storage options,or same-day support contracts typically includedwith standard servers. At the same time, they needhigh-performance hardware with a dense, energyefficientdesign that can help them save on power,cooling, and space. Dell PowerEdge C-Seriesservers can address these requirements—helpingto decrease total cost of ownership with a dense,streamlined design that was purpose-built forscale-out environments.Understanding hyperscalecomputing requirementsHyperscale environments have hardwarerequirements that are distinct from traditionaldata centers. Because these environmentshave focused uses, organizations do not wantto pay for capabilities they do not need. Forexample, because high availability in hyperscaleenvironments is typically achieved primarilythrough software, organizations do not needmany of the redundant hardware componentsor availability tools included with other servers. Inaddition, these organizations do not typically usesame-day parts replacement services; instead,they typically have on-site parts kiosks they canuse to service multiple systems during regularlyscheduled maintenance windows.At the same time, organizations withhyperscale environments want to accelerateaccess to the latest technology. They do not wantto wait for software components to be updatedor hardware components to be validated to gainaccess to the latest processing architecture.Because the environment has a single focuseduse, they do not need broad support for a range ofoperating systems or support for legacy software.14 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Fast accessto high performanceIn designing the Dell PowerEdge C-Series, Dell has built onthe knowledge gained by producing custom servers for largeorganizations running hyperscale environments. For example,the PowerEdge C6100 shares several attributes with acustomized server that the Dell Data Center Solutions (DCS)team created for R Systems, a high-performance computing(HPC) resource provider.When a new customer requested a particular large-scaleHPC cluster for a rapidly approaching product launch,R Systems needed to find a way to accommodate both thenew customer and the company already using that cluster.The R Systems team decided to upgrade a second clusterand migrate the existing customer to that upgraded cluster.R Systems needed dense, energy-efficient servers equippedwith the latest Intel ® architecture, but without the redundanthardware components and software features supplied withexisting PowerEdge models.Dell PowerEdge C6100: 2U designhousing four two-socket server nodeswith shared infrastructure“The Dell DCS group was ableto produce the servers morequickly without having tointegrate some of the typicalsoftware components. Byeliminating redundant powersupplies and other componentsthat are unnecessary in ourparticular deployment model,Dell DCS created a server with anextremely compact form factor.”—Brian KucicVice president of business development at R SystemsOctober 2009R Systems worked with the Dell DCS team to design andproduce a streamlined server that would deliver the requiredperformance, energy efficiency, and density. The serversfeature four nodes in a 2U rack space—a shared infrastructuredesign that was subsequently adopted for the PowerEdgeC6100 server. Each node includes two processors from theIntel Xeon ® processor 5500 series.Because R Systems did not need many of the softwarecomponents included with general-purpose servers, Dellwas able to produce compact servers quickly, withoutrequiring R Systems to wait for software validation on theIntel processors. R Systems was able to accommodate bothcustomers by the launch deadline. The upgraded clusterprovided the existing customer with a 200 percent aggregateperformance increase using half as many nodes as the othercluster, and the dense, energy-efficient servers helpedR Systems control operational costs while conserving datacenter space for continued expansion.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved. dell.com/powersolutions | 2010 Issue 02 15

FeaturesectionSolutions for the virtual eraPowerEdge C1100 PowerEdge C2100 PowerEdge C6100Form factor 1U 2U 2USockets Two Two Two per nodeMemory Up to 18 Double Data Rate 3(DDR3) modulesUp to 18 DDR3 modulesUp to 12 DDR3modules per nodeHard drivesUp to four 3.5-inch or ten2.5-inch disksUp to twelve 3.5-inch disks plustwo internal SSDs (with supportfor up to twenty-four 2.5-inchdisks expected in the near future)Up to twelve 3.5-inch ortwenty-four 2.5-inch disksUp close withthe PowerEdgeC-SeriesPCIe 2.0expansionslotsThree (one x16 slot and twocustom daughtercard slotsproviding SAS and 10 GigabitEthernet support)Four (two x8 slots and twodaughtercard slots providing SASand 10 Gigabit Ethernet support)Two per node (one x16 slot andone x8 mezzanine slot providing internalSAS, internal RAID, 10 GigabitEthernet, and InfiniBand support)In this video walk-through,Dell solutions architect RafaelZamora gives an up-close lookat the first three systems in thePowerEdge C-Series line.youtube.com/watch?v=EcIXX7vj-8sFigure 1. Key features of Dell PowerEdge C-Series serversThese organizations also want to maximizehardware density: IT professionals who arebuilding hyperscale environments are thinkinglarge, but they do not have endless supplies ofpower, cooling, or space. They need to maximizethe number of nodes within the limitations of thephysical infrastructure. (For examples, see the “Fastaccess to high performance” and “Greener cloudcomputing” sidebars in this article.)Introducing Dell PowerEdgeC-Series serversDell PowerEdge C-Series servers were created asthe result of experience gained by the Dell DataCenter Solutions (DCS) team. DCS works withlarge organizations that need customized serversand support for their hyperscale environments.Through these engagements, DCS has learnedfirsthand what types of features and characteristicsenterprises find essential—and which areunnecessary—in a hyperscale infrastructure.PowerEdge C-Series servers help organizationsaddress the requirements of hyperscaleenvironments without customized designs.PowerEdge C-Series servers offer astreamlined approach for targeted hyperscaleenvironments. For these servers, Dell has removedthe redundant hardware, broad OS support,and same-day parts replacement that theseorganizations do not need, helping provide therequisite performance levels in dense, energyefficientconfigurations.These servers also allow organizations to gainfast access to emerging technology, instead ofwaiting for customized solutions or traditionalgeneral-purpose servers (with their additionalfeatures and extensive OS support). IT groups candeploy focused hardware, capitalize on robust Dellservices and support, and even select targetedsolutions that help further simplify deployment.The PowerEdge C-Series comprises threemodels: the PowerEdge C1100, the PowerEdgeC2100, and the PowerEdge C6100. Each canbe equipped with processors from the IntelXeon processor 5500 series or Intel Xeonprocessor 5600 series; a variety of hard drivetypes, speeds, and capacities, including SerialAttached SCSI (SAS) drives, Serial ATA (SATA)16 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Greener cloudcomputingBy working with software-as-a-service (SaaS) providersworldwide, Dell has had firsthand experience providingthese businesses with compact, energy-efficient servers thatfacilitate simplified management. When Swedish SaaS providerNAB Solutions needed to refresh its cloud environment,its administrators decided to implement a virtualizedinfrastructure. For this midsize business, it was imperativeto use a hardware platform that could conserve data centerspace for future growth, minimize administrative complexity,and reduce the costs and environmental impact of IT.NAB Solutions switched to a Dell hardware platformand adopted Dell PowerEdge blade servers for thecloud environment. Each blade includes four multi-coreAMD Opteron processors and supports the large memorycapacity needed for running multiple virtualized serverson each physical host.“Our new Dell environment hascut power consumption by half.Increasingly, companies will onlywork with suppliers who canprove they are truly committedto sustainability. Besides the costsavings, green IT is important inhelping drive new business.”—Jörg WiesemannProject manager and infrastructure specialistat NAB SolutionsJune 2009By creating a virtualized environment with a densehardware platform, NAB Solutions was able to reducethe size of its cloud infrastructure by nearly 87 percent,leaving data center space for future growth. The virtualizedenvironment has also helped simplify key administrativetasks, such as provisioning new servers or movingapplications from one physical server to another duringmaintenance. By using a reduced number of highlyenergy-efficient servers, NAB Solutions has lowered itspower consumption and created an environmentallyfriendly IT environment.drives, and solid-state drives (SSDs); and flexiblePCI Express (PCIe) 2.0 I/O options (see Figure 1).A standard baseboard management controllerwith Intelligent Platform Management Interface(IPMI) 2.0 support helps simplify and streamlineongoing systems management.Each model is suited for different types ofenvironments. The PowerEdge C1100 is designedto deliver outstanding performance and a largememory footprint in a 1U form factor, and can beespecially useful for applications that cache massivequantities of data in memory, including searchengine, Web analytics, and Web 2.0 applications.The PowerEdge C2100 builds on the PowerEdgeC1100 with a 2U design that supports increasedstorage capacity, and is geared toward searchengine, Web analytics, distributed database, andprivate or public cloud computing environments.For maximum density, the PowerEdge C6100offers a shared infrastructure design with fourindependent server nodes in a 2U form factor.Optimized for focused environments withhyperscale application workloads, the PowerEdgeC6100 provides exceptional compute performance,a large memory capacity, and an eco-friendly sharedchassis that helps to reduce the overall weight,power, and cooling requirements for the four nodes.This model is designed to provide twice the densityReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 17

FeaturesectionSolutions for the virtual eraExplore cloud computing withDell IT Consulting servicesThe Dell IT Consulting team can work with organizations that are considering a moveto cloud computing, helping them to understand the potential advantages of buildinga cloud infrastructure and establishing a road map for change.Cloud Workshop: In a one-to-one format, IT Consulting introduces organizations toprivate clouds, public clouds, hybrid models, and related concepts such as software asa service (SaaS), explaining the advantages of adopting a cloud approach.Cloud Assessment: IT Consulting examines current applications and workloads,reviews the current state of the infrastructure, and provides an assessment for cloudcomputing approaches. The consultants then produce a road map with customizedrecommendations and strategies that include integrated cloud hardware and software.Cloud Design and Implementation: IT Consulting streamlines the path to cloudcomputing and helps reduce costs by recommending, procuring, and deploying asuitable combination of technologies for a tailored cloud implementation.of traditional 1U servers and isprogram was designed forthat includes the ability to fast-well suited for many hyperscaleorganizations using PowerEdgetrack dispatch parts and labor,CompellingcloudsPerformance and reliabilityare essential for thesuccessful delivery of cloudcomputing services. Inthis case study, read howSwedish company CityNetwork Hosting deployedvirtualized Dell PowerEdgeblade servers and DellEqualLogic storage to createan efficient, reliable, scalableplatform for its new cloudoffering.dell.com/downloads/global/casestudies/2010-city-network-10008302.pdfenvironments, includingHPC clusters. In addition,the PowerEdge C6100 helpssimplify service: unlike somecomparable shared infrastructureservers from other vendors, allnodes in the PowerEdge C6100are hot-serviceable, enablingadministrators to power downany one node without affectingthe availability of the other nodesin the system.Tailoring supportwith Dell ServicesOrganizations can capitalize onC-Series servers, and helpsaccelerate deployment ofhyperscale infrastructures byracking, stacking, and cablingservers before shipping them.Organizations receive fullyassembled racks that are readyto be powered up. This servicealso provides rack configurationdocuments with configurationdetails that help significantlyreduce implementation time.Organizations can alsoselect from a wide range of DellSupport Services options to helpstandardize the maintenancebypassing basic troubleshooting.Enterprise-Wide Contractis an enterprise-level supportservice that provides a designatedservices delivery manager whofacilitates proactive planning andreporting to help organizationsmaximize the uptime andperformance of their Dellinfrastructures. Organizationscan also select self-maintenanceoptions geared toward largescaleenterprises, includingon-site parts kiosks for immediateaccess to replacement parts,periodic on-site maintenanceservices developed specificallyof their Dell infrastructures. Theservice, and the Dell Online Selffor PowerEdge C-Series serversBasic Support option providesDispatch program. As part of theas well as many of the sameentry-level hardware supportOnline Self Dispatch program,support options availableand includes call-in assistanceIT administrators can earn afor standard PowerEdgeduring business hours only. Dellcertification to order warrantyservers. For example, the DellProSupport for IT is availablereplacement parts directly fromConfiguration and Deploymentfor 24/7 professional-levelDell. Dell IT Consulting also offersRack Integration Serviceshardware and software supporta comprehensive set of services18 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

to help organizations understand and implementcloud computing in their own environments(see the “Explore cloud computing with Dell ITConsulting services” sidebar in this article).Accelerating deploymentwith Dell Cloud Computing SolutionsThe upcoming Dell Cloud Solution for WebApplications is expected to be the first turnkeycloud offering that combines software withan optimized architecture powered by DellPowerEdge servers and a broad range of supportservices. This solution is designed for organizationsdeploying Web applications on a private,on-premise cloud as well as telecommunicationsand hosting organizations looking to buildinfrastructures for a public cloud. When available,it is expected to deliver the following:• Enhanced volatile Web traffic management:Load balancing and processor-burstingcapabilities allow applications to scale as userdemand fluctuates, helping maintain rapidresponse times.• Purpose-built support for Web applications:Built-in support for applications written inJava, PHP, Python, Ruby on Rails, and otherprogramming languages helps rapidly delivervirtualized instances that are ready to run Webapplications such as Apache HTTP Server orApache Tomcat.• Support for lab to hyperscale deployments:Organizations can deploy simple lab clouds fordeveloper workgroups as well as hyperscaleproduction clouds that can scale to hundredsor thousands of virtualized compute instances.• Self-service portal: End users can use a selfserviceportal to help simplify the process ofacquiring compute resources to begin newprojects or launch new applications. The portalprovides access to compute resources and helpsIT departments respond to the needs of theorganization as a whole.• Physical infrastructure: Supportedhardware platforms include PowerEdge C2100servers or PowerEdge R710 servers plusDell PowerConnect switches and DellPowerVault storage.• Services: Dell IT Consulting, Configuration andDeployment, and Support Services options havebeen developed to help organizations plan,implement, and maintain the Dell Cloud Solutionfor Web Applications.By combining Joyent cloud software withDell hardware and services plus an optimalblueprint for creating a private cloud, thissolution helps organizations get started rapidlyand operate the infrastructure efficiently withhighly responsive, on-demand scaling.Reducing total cost of ownershipUsed as a building block for hyperscale serverenvironments, Dell PowerEdge C-Series servershelp organizations reduce their total cost ofownership. These servers avoid hardware,software, and support components that areunnecessary in hyperscale deployments, providinga dense, energy-efficient design that packs atremendous amount of processing performanceand memory capacity while helping to reducepower, cooling, and space requirementsalong with their associated costs. The cloudinfrastructure itself is designed to keep costsdown by maximizing the utilization of hardwareresources. A wide range of service options helporganizations simplify deployment and ongoingmaintenance so that they can begin to realize thebenefits of cloud computing quickly and easily.Learn moreSteven Croce is the product marketing managerfor the Dell PowerEdge C1100 and PowerEdgeC2100 servers.Brandon Draeger is the product marketingmanager for the Dell PowerEdge C6100 serverand the Dell Cloud Solution for Web Applications.Buck Avey is the services product planningmanager for Dell PowerEdge C-Series servers andthe Dell Cloud Solution for Web Applications.Dell PowerEdge C-Series servers:dell.com/poweredgecDell cloud computing:dell.com/cloudReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 19

FeaturesectionSolutions for the virtual eraAdvancing scalabilityand performancewith four-socket DellPowerEdge serversBy Armando Acosta and Robert BradfieldFour-socket Dell PowerEdge R810 and PowerEdge R910rack servers and PowerEdge M910 blade servers offer highlyscalable processing power and memory capacity to helpcost-effectively boost performance and expand virtualizedenvironments, while other features help increase reliabilityand simplify administration.Enterprises running large-scale databasesor virtualized environments need serversthat let them scale processing powerand memory capacity flexibly andcost-effectively. Because servers in these typesof environments run everything from enterpriseproductivity tools to mission-critical businessapplications, they must provide outstandingreliability. And to help optimize worker productivityand enhance business agility, servers must alsobe quick to deploy, simple to manage, and easyto maintain. The Dell PowerEdge R810 andPowerEdge R910 rack servers and PowerEdgeM910 blade servers are designed to meet theserequirements, combining scalable processingperformance and tremendous memory capacitywith a dense server design, reliability features, andstreamlined management.Achieving scalablecompute performanceDell PowerEdge R810 and PowerEdge R910 rackservers and PowerEdge M910 blade servers aredesigned to provide exceptional scalability withineach server, helping organizations to quickly andcost-effectively expand database environmentsor add virtual machines to each physical host.In the past, scaling processor performancefrequently meant adding servers. With thePowerEdge R810 rack server and PowerEdgeM910 blade server, however, organizations canbegin with a two-socket configuration and thenscale to four sockets when they need to increaseapplication performance or extend a virtualizedenvironment (see Figure 1). Organizations cantherefore double the processing performancewithin a server without needing to provide20 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Dell PowerEdge R810rack serverSocketsForm factorProcessorsDell PowerEdge M910blade serverDell PowerEdge R910rack serverTwo or four Two or four Four2U rack server Full-height blade server 4U rack serverIntel Xeon processor 7500or 6500 seriesIntel Xeon processor 7500or 6500 seriesIntel Xeon processor7500 seriesMaximumperformanceForrest Norrod, vice presidentof the Dell Server Group,answers questions on thelatest Intel Xeon processor–based Dell servers and howthey can help IT organizationsincrease performance,reliability, and efficiency formission-critical applications.youtube.com/watch?v=Y7OCYT9WSmsMemory slots32 32 64Maximum memory capacity512 GB 512 GB 1 TBPCI Express (PCIe) slots6 PCIe slots Up to 4 mezzanine card slots Up to 10 PCIe slotsFigure 1. Four-socket Dell PowerEdge rack and blade server configurationsthe extra power, cooling, ordata center space associatedwith additional servers.Each of the four sockets inthese PowerEdge servers can befilled with powerful multi-coreIntel ® Xeon ® processors thatoffer significant performanceenhancements compared withprevious-generation PowerEdgeservers. The PowerEdge R810rack server and PowerEdgeM910 blade server can beequipped with the Intel Xeonprocessor 7500 or Intel Xeonprocessor 6500 series; thePowerEdge R910 rack servercan be equipped with theIntel Xeon processor 7500series. Using this generationof Intel Xeon processors, thePowerEdge R910 rack serverhas achieved up to 219 percenthigher performance comparedwith the previous-generationPowerEdge R900 rack server,and the PowerEdge M910 bladeserver has achieved up to 76percent higher performancecompared with the PowerEdgeM710 blade server in a foursocketconfiguration. 1By using the latest multicoreIntel Xeon processors,these servers can dramaticallyincrease the number ofprocessor cores per rackcompared with previousgenerationPowerEdge servers.In its four-socket configuration,the PowerEdge R810 servercan be equipped with 32processor cores in a single 2Uenclosure—up from a maximumof 12 cores in previousgeneration2U PowerEdgeservers. Both the PowerEdgeR910 4U rack server and the1 Based on SPECjbb2005 performance testing by Dell Labs in March 2010; actual performance will vary based on configuration, usage, and manufacturingvariability. For the latest SPECjbb2005 benchmark results, visit spec.org.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 21

FeaturesectionSolutions for the virtual era219% 76% 1/3The Dell PowerEdge R910rack server has achievedup to 219 percent higherperformance than thePowerEdge R900.The Dell PowerEdge M910blade server has achievedup to 76 percent higherperformance than thePowerEdge M710.The Dell PowerEdge M910blade server providesdouble the memorycapacity of previousgeneration4U Dell serversin one-third the space.PowerEdge M910 full-heightblade server also support up to32 cores per server, comparedwith a maximum of 24 coresin the PowerEdge R900 rackserver and 12 cores in thePowerEdge M710 blade server.In addition, the thread-levelparallelism provided by IntelHyper-Threading Technologyenables highly efficientuse of these processingcores—providing a substantialperformance boost for multithreadedsoftware comparedwith the same number of coreswithout this technology.Packing additional coresinto each server enablesorganizations to support largedatabases and large-scalevirtualized infrastructures whilehelping to keep power, cooling,and space requirements undercontrol. For example, thePowerEdge M910 server enablesthe deployment of up to eightfour-socket blade servers in just10U of rack space. Comparedwith previous-generation 4UDell servers, the PowerEdgeM910 server is designed toprovide double the memorycapacity in one-third the space.The powerful processingcapabilities and compact designsof these servers contributes toexceptional price/performanceratios. At the same time,supporting additional processorsin a reduced amount of spacehelps organizations minimizepower consumption. In foursocketconfigurations, thesePowerEdge servers also deliveroutstanding performance perwatt compared with previousgenerationPowerEdge servers.Breaking the servermemory capacity barrierMany organizations runninglarge enterprise applicationsand virtualized environmentsfind that server memorycapacity is a primary limitationfor scaling. These four-socketDell PowerEdge serversaccommodate significantlygreater memory capacity thanprevious-generation PowerEdgeservers, enabling organizations tobreak through those limitations.The PowerEdge R810 andPowerEdge M910 servers bothoffer 32 memory slots for upto 512 GB of memory in eachserver; for the PowerEdge M910server, that is more than twicethe maximum memory capacityof previous PowerEdge bladeservers. The PowerEdge R910server provides 64 memoryslots for up to 1 TB of memory.These large memorycapacities enable organizationsto increase the number of virtualmachines per server or enhancethe performance of memorybounddatabase applications.They can also capitalize onthe large number of memoryslots to help reduce costs:populating these slots withsmaller, less-expensive memorymodules enables organizationsto provision sufficient memorywithout incurring the cost oflarger-capacity modules.In addition, Dell FlexMemBridge technology allowsorganizations to take advantageof large memory capacitieswithout having to scale thenumber of processors. Thisinnovative, patent-pending22 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

FeaturesectionSolutions for the virtual eraIntroducing theDell PowerEdge R815:High performance in acompact designMaximizing performance while making efficient use ofavailable power and space can be critical in data centerenvironments. The Dell PowerEdge R815 rack serverwith the AMD Opteron 6000 Series platform deliversfour-processor performance in a compact, cost-effective2U design.Enterprise IT departments areconstantly looking for ways to increaseperformance, memory capacity,and I/O bandwidth while extractingmaximum value from the available space. Forthese environments, the powerful Dell PowerEdgeR815 rack server offers an innovative combinationof high performance, compact size, and costeffectivevalue. Enabled by the high core countand substantial memory capacity of the AMDOpteron 6000 Series platform, the PowerEdgeR815 is a powerful x86-based four-processorserver in a 2U form factor—helping organizationsto achieve high levels of processing density,flexibility, and energy efficiency at a price typicallyassociated with two-processor servers.Increased density and flexibilityTraditionally, x86-based four-processor rackservers have been available primarily in a 4Uform factor, meaning that organizations thatneeded these levels of processing and memoryresources could fit no more than 10 suchservers in a typical 42U rack. Thanks in part tothe ample core count and memory footprintof the AMD Opteron 6000 Series platform,however, the PowerEdge R815 can deliver fourprocessorperformance in a 2U design—allowingorganizations to install up to 21 of these servers ina 42U rack, and enabling them to implement thecapacity of a four-processor system while morethan doubling their rack density. When takinginto account the costs of data center space,the savings from consolidating servers onto areduced number of racks can be considerable.In terms of compute density, a rack holding10 four-processor, 4U servers with 6 cores perprocessor would contain just 240 total cores.Because each four-processor server based on theAMD Opteron 6000 Series platform can include24 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

“Like other 11th-generationPowerEdge servers, thePowerEdge R815 is designedfor long-term value.”The Dell PowerEdge R815 with the AMD Opteron 6000 Series platformpacks four-processor performance into a 2U rack server designup to 48 cores, a 42U rackholding 21 PowerEdge R815servers could contain up to1,008 total cores—a 320 percentincrease over legacy 4U servers.This compute density even rivalsthat of blade servers, enablingorganizations to choose amonghighly flexible deploymentoptions to help meet theirspecific needs.The flexibility of thePowerEdge R815 also extendsto virtualized environments.PowerEdge servers and theAMD Opteron 6000 Seriesplatform both support a widerange of hypervisors, includingMicrosoft ® , VMware ® , andCitrix ® platforms. As a result,organizations using thePowerEdge R815 as a virtualizedhost server can select theappropriate hypervisor for theirspecific needs.Enhanced energy efficiencyIn addition to rack and floorspace, the PowerEdge R815 isdesigned to conserve poweras well. First, the AMD-Psuite of power managementtechnologies built into AMDOpteron processors enablesthe AMD Opteron 6000Series platform to deliver12-core performance withinthe same general powerand thermal envelopes asprevious-generation 6-coreprocessors. And second,because the PowerEdge R815is a 2U server, it contains justtwo power supplies insteadof the four that are typical ofsome 4U four-processor servers,helping reduce power draw.The PowerEdge R815also incorporates key DellEnergy Smart technologiesto enhance efficiency.The careful positioning ofsystem components andinternal shrouding, alongwith programmable voltageregulators, efficient fandesigns, and system thermalmanagement, is designed toprovide the right level of coolingto help reduce energy use.Highlighting thePowerEdge R815In this video, Dell seniorproduct line consultantArmando Acosta outlinessome of the key features ofthe PowerEdge R815 rackserver and the advantagesit can provide in enterpriseIT environments.youtube.com/watch?v=qOl88iZTu2gReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 25

FeaturesectionSolutions for the virtual era1,008A 42U rack filled with Dell PowerEdgeR815 servers can house up to 1,008total processor cores—a 320 percentincrease in compute density overlegacy 4U servers.Sensors also help decrease power consumptionby automatically tracking and regulating internalthermal activity. And Energy Smart managementfeatures such as power capping, powerscheduling, and device disablement further helpPowerEdge R815 servers to limit energy use.Intelligent management capabilitiesLike other 11th-generation PowerEdgeservers, the PowerEdge R815 incorporates avariety of sophisticated management featuresdesigned to simplify administration and reduceoperating costs, including the following:• Lifecycle Controller: The Lifecycle Controlleris an embedded chip installed directly on themotherboard that makes common provisioningfunctions such as deployment, configuration,and updating available to administratorswithout requiring media such as a CD or DVD.By enabling administrators to perform thesetasks without the hassle of tracking and usingseparate media, this chip helps to streamlineserver configuration and reduce downtime.• Interactive LCD screen: Positioned on thefront of each server, this miniature monitorenables technicians to see error messages,execute selected boot options, view powerconsumption information, and more.• Dell Management Console: A standardfeature on Dell servers, Dell ManagementConsole powered by Altiris from Symantec offers IT staff a consolidated view of theirinfrastructures, including deployment,inventory, monitoring, and updates.Also like other 11th-generation PowerEdgeservers, the PowerEdge R815 is designedfor long-term value. Dell incorporatesdurable materials, emphasizes componentredundancy and ease of servicing, andperforms rigorous testing and validation to helpminimize downtime and ensure reliability.High-performance,highly efficient rack serverBalancing performance against power, space,and cost requirements can be a challenge forenterprise IT departments. Designed to supporthighly threaded workloads such as virtualization andhigh-performance computing, the PowerEdge R815server with the AMD Opteron 6000 Series platformprovides a highly efficient way for organizationsto gain the performance advantages of a fourprocessorserver while still maintaining the spaceand cost efficiencies of a 2U design.Learn moreDell PowerEdge servers withAMD Opteron processors:dell.com/poweredge/amd26 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

www.f5.comdellteam@f5.com

FeaturesectionSolutions for the virtual eraHow deduplication helpsreduce the cost of backup anddisaster recoveryBy Kay Benaroch and Shane JacksonDell and EMC have teamed up to deliver efficientdisk-based backup and recovery with integrateddeduplication technology designed to make disk-basedbackups as cost-effective as tape-based backups.Deduplication pays offDell/EMC DD Series appliancesincorporate deduplication and other keytechnologies to help meet storage andbackup challenges.• Up to 99 percent bandwidth reductionfor replication• 90–97 percent data reduction forbackups• Leading defense against data integrityissues with the Data Domain DataInvulnerability ArchitectureToday’s explosive data growth is prompting manyorganizations to look for ways to increase the efficiencyof their storage, backup, and disaster recovery processes.The growth of data is straining capacity, but at the sametime, cost and complexity frustrate many IT managers seeking tomake improvements. For example, tape-based approaches thatworked well when originally implemented can prove to be slow,expensive, and unreliable as the organization grows and as tapesmust be physically transported from remote sites to a centralstorage location. Disk-based backup is inherently faster and morereliable than tape, but has traditionally been more costly, and largedisk arrays can be difficult to manage.Dell and EMC are addressing these challenges by leveraging thedata reduction and cost-saving advantages of deduplication storagesystems from Data Domain, which was purchased by EMC in 2009.Rather than adding deduplication to an existing storage platform, Delland EMC have introduced the Dell/EMC DD Series—purpose-builtappliances that incorporate disk-based storage and deduplication.These appliances are designed to tame data growth as effectively aspossible, helping organizations to reduce complexity and costs.Overcoming challenges to effective data protectionCost pressures can keep many organizations from making neededstorage and backup improvements to cover the variety of operatingsystems, applications, and geographic locations that produce critical28 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

BackupData center applications(Microsoft, Oracle, SAP, VMware)ArchiveNAS, SAN, DASPrimarystorageFigure 1. Dell/EMC DD Seriesinfrastructure and ecosystemBackup applications(EMC, CommVault,Symantec, Vizioncore, CA,IBM, Hewlett-Packard,Atempo, BakBone)Archive applications(EMC, CommVault,Symantec, F5)NetworkReplicationover WANWANDisaster recoveryDell/EMC DD SeriesapplianceDell/EMC DD Seriesapplianceinformation. IT managers may be forcedto deploy systems piecemeal, requiringadditional management by IT departmentsthat are already stretched, or they mayhave resources for headquarters or centralsites but lack the budget to create acomprehensive plan that encompasses alltypes of servers and applications at local andremote sites.Organizations may also be exposed torisk by doing without essential protectionsuch as off-site disaster recovery. Off-siteprocesses for disaster recovery havehistorically induced latency, promptingmany organizations to simply createlocal copies to speed local recovery—anapproach that exposes the organizationto data loss in the event of accidentaldeletion or site disasters. In addition, thecost of bandwidth to replicate the growingvolume of critical information over a widearea network (WAN) is often prohibitive,especially for remote or branch offices.Complexity in the backup environmentis another factor that can delay storage andbackup improvements. Many organizationshave multiple backup software instancesor heterogeneous environments. Thiscomplexity can increase operational andprocedural errors, leading to unprotected,corrupted, or lost backup data. As a result, ITmanagers may not be able to meet businessneeds for data restoration and recovery.To address these challenges withoutrequiring a costly redesign of backupprocedures, IT managers are exploring waysto remove duplicate data from the backupprocess, reducing data volume to the pointthat they can use fewer hard drives thanthey had been. Finding an approach thatworks with a flexible range of backup andarchive software is also desirable. Dell andEMC are working collaboratively to createstorage and backup systems that includeintegrated deduplication functionality.Examining technical requirementsBuilding a storage system that delivers the fullpotential of deduplication requires a flexibleapproach that addresses variable-lengthduplicates, multiple formats and protocols,and other technical considerations.Variable-length duplicates. Conventionalstorage systems, whether network attachedstorage (NAS) or storage area networks(SANs), store data in fixed-size blocks.Some deduplication systems simply lookfor identical fixed-size data blocks, anapproach that is inadequate for maximizingdeduplication results; to be effective, asystem must be able to identify variablelengthsegments as well. This capabilityis necessary to maximize the amountof redundant data that can be found inincoming data blocks, regardless of smallchanges in those blocks compared withprevious backups.Format parsing. Data comes in manyformats generated by many differentapplications, and the same duplicate data isoften embedded in those different formats. Thesheer number of these formats and the speedat which they change make it impractical fora storage vendor to support them all. Parsingthe formats requires substantial overhead. AReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 29

FeaturesectionSolutions for the virtual eraEnhancing logical capacityComparing physical capacity to logical capacity illustrates the value of integrateddeduplication in Dell/EMC DD Series appliances. The physical capacity is theamount of raw storage provided by the disks in a particular system. Some ofthis storage is consumed by appliance operations such as RAID and spares;subtracting this amount from the physical capacity yields the usable capacity,to which the expected deduplication ratio is then applied. Multiplying usablecapacity by the deduplication ratio yields the expected logical capacity—theamount of backup data an administrator can keep on the system afterdeduplication, depending on change rate, retention policies, and other factors.For example, 1.5 TB of physical capacity in the Dell/EMC DD140 applianceultimately provides up to 17.0 TB of logical capacity with a typical 20:1deduplication ratio (see Figure A). That increase can help deliver substantialadditional terabytes of logical backup capacity to the organization, allowingadministrators to increase the scope of information backed up for replicationand disaster recovery purposes.Physical capacity Logical capacity Maximum throughputDell/EMC DD140 1.5 TB Up to 17.0 TB Up to 450 GB/hourDell/EMC DD610 Up to 6.0 TB Up to 75.0 TB Up to 675 GB/hourDell/EMC DD630 Up to 12.0 TB Up to 165.0 TB Up to 1.1 TB/hourFigure A. Capacities and throughputs for Dell/EMC DD Series appliancesCalculatingthe benefitsBased on a few simplequestions, the Data DomainDeduplication Calculator canhelp organizations evaluatehow deduplication couldhelp reduce storage andbandwidth requirements intheir specific environments.dedupecalculator.comstorage-based deduplication engine should be dataagnostic, and find and remove duplicates in data nomatter how it is packaged or stored in the system.Multiple protocols. Many standard accessprotocols are used in storage systems today, fromCommon Internet File System (CIFS) and NetworkFile System (NFS) to block-based and virtualtape library (VTL) access methods. For example,user directories may be in NFS, the Microsoft ®Exchange server may need to run in data blocks,and backups may require VTL. To be efficient, abackup storage system should support all of theseprotocols, and the deduplication approach shouldbe able to remove redundant data no matter howit is stored.Processor-centric versus disk-intensivealgorithms. Over the last two decades, processorperformance has increased dramatically comparedwith disk performance. Today, processorperformance takes another leap with every doublingof the number of cores in a chip. Algorithmsdeveloped today for deduplication should takeadvantage of the growth in processor performanceinstead of being tied to disk performance.Deduplicated replication. When it comes todisaster recovery, true data protection requiresstoring a copy of the data safely at a remote location.Replication has long been used for the relatively smallvolume of mission-critical or high-value data in anorganization, but many organizations find that thecost of replication can be too high for the remainderof their information.Backup storage systems designed with integrateddeduplication and replication can help reduce thebandwidth required to replicate large quantitiesof data from one site to another, thus enhancingthe cost-effectiveness of the replication process.However, not all deduplication systems canreplicate, and even those that can have wide gapsin capabilities. The system must be implementedin a way that runs fast enough for deduplicationand replication with low overhead across a30 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

comprehensive range of topologies and infrastructuretypically found in the distributed enterprise.Combining disk-based backupwith integrated deduplicationIn an effort to bring the cost of disk-based backupclose to that of tape-based backup, Dell has beenincorporating deduplication into an increasingnumber of storage technologies. The Dell/EMC DDSeries of storage appliances increases the breadthof options for disk-based backup with deduplication.The Dell/EMC DD Series consists of threemodels: the Dell/EMC DD140, the Dell/EMCDD610, and the Dell/EMC DD630. Each offersdifferent levels of capacity, throughput, andscalability. (For capacity and throughput details,see the “Enhancing logical capacity” sidebar in thisarticle.) Each self-contained appliance has its ownconnectivity and power cable, so it can simply pluginto an existing data center environment.Dell/EMC DD Series storage appliancesincorporate leading Data Domain disk-baseddeduplication storage systems. Data Domainsystems are “CPU-centric” and therefore are notdependent on disk drive performance advances oran increase in the number of disk drives to speedthroughput over time. They are also designed toidentify duplicate data regardless of how the data ispackaged and regardless of changes in block size.Because these appliances are built fordeduplication from the ground up, they can run fastenough to apply the benefits of deduplication toreplicated data streams in real time, so that the laguntil data is safely off-site can be as small as possible.Systems that have add-on deduplication can imposeunnecessary delays on the replication process.Assessing Dell/EMC DD Series backupand recovery featuresDeploying disk-based backup using Dell/EMC DDSeries appliances with integrated deduplication canprovide benefits that address many IT concernsregarding backup and recovery and help providefavorable return on investment.Complexity reduction. Dell/EMC DDSeries appliances provide a deduplicationapproach that can be implemented easily inheterogeneous environments (see Figure 1).They can connect to storage using CIFS and NFSover Ethernet, Symantec OpenStorage overEthernet, and VTL over Fibre Channel, and canwork with many common backup and archivesoftware applications—enabling organizations toconsolidate backups but maintain current datamanagement schemes. By simplifying backupand recovery processes, organizations can savetime and money while freeing staff resources topursue other tasks that help increase productivityand profitability.Data reduction. Dell/EMC DD Seriesappliances with integrated deduplication aredesigned to avoid the storage of redundantdata, helping reduce the disk capacity requiredfor backup. Reducing the number of hard drivesneeded to store data then helps reduce the costsof power, cooling, space, and management inthe data center. Space that was allocated tothe physical footprint for extra drives can beavailable for other uses. Using Dell/EMC DD Seriesdeduplication technology, organizations canrealize up to a 99 percent bandwidth reduction forreplication in a typical backup environment witha 5–10 percent incremental change rate, and canachieve 90–97 percent backup data reduction witha typical 10:1–30:1 deduplication ratio.Low overhead and high speed. BecauseDell/EMC DD Series appliances use in-linededuplication, they do not require the overheadof extra disk capacity found in post-processingapplications of deduplication. Performance is notdependent on adding disk drives, and equipment,management, and operating costs can be reducedcompared with other approaches to storagebackup. Deploying Dell/EMC DD Series appliancesalso helps speed recovery times beyond whattape-based backup can deliver, helping reduce theimpact to an organization if something goes wrong.Cost-effective disaster recovery. Replicationto a remote disaster recovery site can behighly cost-effective with Dell/EMC DD Seriesappliances, because deduplication reduces theamount of data that needs to be transmittedand the required bandwidth. Bandwidth is oneof the major expenses involved in replication,and by helping reduce bandwidth requirements,deduplication can make replication a viableTaming theunstructureddata beastThis blog post outlines threeprimary ways that organizationscan address out-of-control datagrowth, including deduplicationwith Dell/EMC DD Seriesstorage appliances.en.community.dell.com/dell-blogs/b/direct2dell/archive/2010/03/24/taming-the-unstructured-data-beast.aspxReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 31

FeaturesectionSolutions for the virtual era99%disaster recovery method for a wide range of datasets and applications, and for many organizationsthat may have previously ruled out robustprotection as a cost-prohibitive option.Data integrity and protection. The DataDomain Data Invulnerability Architecture inDell/EMC DD Series appliances is designed toensure data integrity and protection throughcontinuous fault detection and healing, end-toendverification of data recoverability at time ofbackup, redundant system components, andother data safety techniques. Unlike some otherdeduplication systems or file-based backupsystems, Dell/EMC DD Series appliances helpensure that recoverability is verified and thencontinuously re-verified.Tape avoidance. By enabling organizationsto avoid redundant data, Dell/EMC DD Seriesappliances with integrated deduplication help toeliminate space-consuming tape libraries. Theseappliances are designed to deliver all the benefitsof backup to disk, including accelerated backupand restore times and enhanced reliability anddata integrity compared with tape. More datacan be stored with reduced media resourcesDell/EMC DD Series deduplication technologycan help organizations reduce bandwidth forreplication by up to 99 percent.compared with tape, helping to speed therecovery process.Optimizing the storage infrastructureDeduplication has the potential to reduce databackup volume, enhance data protection, and,in general, simplify data management acrossthe storage infrastructure. Dell/EMC DD Seriesappliances offer a cost-effective deduplicationand storage backup option that enhances dataprotection and accelerates recovery time. Theappliances implement deduplication in a way thatcombines performance and low overhead acrossa comprehensive range of backup environments.IT managers can easily deploy mature, flexible,high-throughput deduplication without needing tochange current backup processes.Dell/EMC DD Series appliances offerorganizations a cost-effective storage, backup,and recovery approach designed to reducecomplexity through a single point of contact atDell. In addition, Dell has a suite of fixed-durationprofessional services that help organizationsidentify the optimal approach to archiving, backup,and deduplication to meet their specific needs.Kay Benaroch is a seniormarketing consultantat Dell focused on dataprotection, deduplication,and object-orientedstorage, and hasmore than 30 years ofexperience in marketingtechnology products.Shane Jackson is seniordirector of productmarketing for the EMCBackup RecoverySystems division, andhas more than 17 yearsof experience in thestorage industry.Learn moreDell/EMC DD Series:dell.com/emcDell storage solutions:dell.com/storageReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.nGenius ® InfiniStream ® 2900 Series ApplianceMore Visibility in More PlacesExtended “always-on” deep packet capture into more areas of your network isnow available with the new, compact nGenius® InfiniStream® 2900 Seriesappliance. Enhance your on-demand troubleshooting and end-to-end serviceassurance capabilities.© 2010 NetScout Systems, Inc. All rights reserved.Network Forensic Analysis• Deep Packet Capture• Flexible Data Mining• Application Intelligence• Sniffer Decodes and Experts2906/GS: 2-Port 10/100/1000Base-T, 500GB ........A28629492910/GS: 4-Port Gigabit Configurable, 500GB .......A29576352916/GS: 4-Port 10/100/1000Base-T, 500GB ........A2883635For more information contact yourDell account team or visit Dell.com50100465-netscout.indd 16/3/10 10:48 AM

Announcing APC’s new, interactiveenergy-saving Smart-UPS.Intuitive alphanumeric display:Get detailed UPS and power qualityinformation at a glance – including status,about, and diagnostic log menus in yourchoice of up to five languages.Energy savings:A patent-pending “green” mode achievesonline efficiencies approaching 99 percent,reducing heat loss and utility costs.Configurable interface:Set up and control key UPS parametersand functions using the intuitive navigationkeys. On rack/tower convertible models, thedisplay rotates 90 degrees for easy viewing.Contact yourDell Representativetoday and ask forAPC Smart-UPSIf you want Legendary Reliability inside, it had better say APC outside.What do you get when you combine 25 years of Legendary Reliability with the latest in UPS technology?Introducing the new APC Smart-UPS range of interactive, intuitive, and energy-saving UPSs, designed toprotect critical server and network equipment from power threats and downtime.New APC Smart-UPS: Smarter. Easier. Greener.Thanks to millions of dollars in research, APC can proudly claim that only the new Smart-UPS features theunique battery life expectancy predictor, telling you the exact month and year for battery replacement. Precisiontemperature-compensated charging extends battery life; unique power meter function monitors energy usage;and a patent-pending “green” mode boosts online efficiencies up to 99 percent, saving on utility costs. Plus, theinteractive LCD provides detailed status, configuration, and diagnostic information previously available only viasoftware.When dollars count and performance is critical, insist on the more intelligent, more intuitive APC Smart-UPS.Now more than ever, the name on the outside guarantees reliability on the inside: APC Smart-UPS.Only APC offers the mosttechnologically advanced,user-friendly features, and theguaranteed reliability you needto protect your critical data andequipment. Look for APC on theoutside to ensure LegendaryReliability on the inside.Download a FREE copy of APC White Paper #10, “PreventingData Corruption in the Event of an Extended Power Outage.”Visit www.apc.com/promo Key Code t441wCall 888-289-APCC x0000 • Fax 401-788-2797©2010 Schneider Electric Industries SAS, All Rights Reserved. Schneider Electric, APC, Smart-UPS, and Legendary Reliability are owned by Schneider Electric, or its affiliated companies in the United States andother countries. e-mail: esupport@apc.com • 132 Fairgrounds Road, West Kingston, RI 02892 USA • 998-2158

FeaturesectionSolutions for the virtual eraOptimizing managementefficiency withDell/EMC NS storageBy Annette Cormier, Eric Cannell, and Brad BunceThe Dell/EMC NS family of unified storage enablesorganizations to easily and cost-effectively consolidate bothunstructured file-system data and structured application datainto a single shared storage system—helping organizationscontrol costs, streamline management, and enhance storageavailability and performance.Explosive growth in content, especiallyunstructured file-system content,can place a heavy burden on storageadministrators in many organizations.Often, they deploy multiple storage systems tohandle file-system-level and block-level storage,including network attached storage (NAS) as wellas Internet SCSI (iSCSI) and Fibre Channel storagearea network (SAN) arrays.However, deploying and managing multiplesystems can be expensive and complicated.Provisioning these systems can be costly, andmanaging them can require significant timeand expertise—increasing the risk of data loss,poor performance, and downtime. Additionally,organizations often overprovision storage tomeet service-level agreements, which can furtherincrease storage costs.An emerging approach to the challenge ofmanaging multiple, disparate storage platforms isunified storage. Using a unified storage approach,both structured and unstructured data can bestored on a single shared system, helping avoidthe need to purchase, deploy, and managemultiple storage platforms. However, for a unifiedstorage system to be effective, it must deliverthe performance, availability, and flexibility thatorganizations have come to expect from purposebuiltstorage such as Fibre Channel SANs.To help organizations enhance the efficiencyand manageability of their enterprise storageenvironments, Dell recently introduced theDell/EMC NS family of unified storage systems.Dell/EMC NS storage arrays support integratedCommon Internet File System (CIFS), NetworkFile System (NFS), iSCSI, and Fibre Channelconnectivity, and are designed for high levels ofperformance, flexibility, and availability. Thesearrays are also available with a wide range offeatures designed to streamline management,protect data, and optimize efficiency, capacity,and cost across mixed Microsoft ® Windows ® ,Linux ® , and UNIX ® environments.Consolidating data and file systemsand scaling storage capacityThe Dell/EMC NS family of unified storagesystems enables organizations to consolidate filesystemand application data into a single sharedstorage environment. In particular, Dell/EMC NS34 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

iSCSI SANNASFibre Channel SANEthernetfile sharingFigure 1. Integrated protocolsupport in Dell/EMC NS storageEthernetiSCSI SANFibre ChannelSANDell/EMC NS unified storagestorage combines support for a range of industrystandardprotocols—including CIFS in MicrosoftWindows environments and NFS in Linux and UNIXenvironments, as well as iSCSI and Fibre ChannelSAN connectivity protocols—that enable integrationof both NAS file-based storage and SAN block-basedstorage (see Figure 1).By supporting multiple file-system protocols,Dell/EMC NS storage can enable consolidation andfile sharing across heterogeneous Windows, Linux,and UNIX environments. Dell/EMC NS storagealso takes advantage of native file-system protocolfeatures. For example, the CIFS protocol leveragesstorage-based file-locking capabilities required inenterprise database environments, and the NFSprotocol delegates file and volume management tothe NAS server to optimize host cycles for missioncriticaloperations. Dell/EMC NS storage can also bewell suited for organizations that need to supportWindows, Linux, or UNIX file-system storage as wellas application storage for VMware ® virtualization,Oracle ® database, Microsoft SQL Server ® database,Microsoft Exchange messaging, and other software.In addition to supporting multiple connectivityprotocols, Dell/EMC NS storage arrays also supporta wide range of drive types to help provide targeted,cost-effective performance, including enterpriseflash drives (EFDs), Fibre Channel drives, and SerialATA (SATA) drives. EFDs offer extremely highperformance,low-latency storage that is well suitedto highly performance-sensitive applications; FibreChannel drives can offer reliable, high-performancestorage; and energy-efficient, low-power SATA IIdrives can offer cost-effective, high-capacity storagefor applications that are not performance intensive,such as archiving or historical data storage.Dell/EMC NS storage arrays are also scalable, andcan be easily expanded to support growth in capacityand throughput requirements (see Figure 2). TheDell/EMC NS-120 array supports a maximum of120 drives, while the Dell/EMC NS-480 array supportsa maximum of 480 drives. The Dell/EMC NS-960array, expected to be available later this year, isdesigned to support up to 960 drives. These drivecounts can be utilized for block storage, file storage, orboth simultaneously based on specific requirements.X-Blades(data movers)ControlstationsStorageprocessorsMaximum numberof drivesMaximum filestorage capacityDell/EMC NS-120 1 or 2 1 2 120 across 8 enclosures 64 TB (32 TB per X-Blade)Dell/EMC NS-480 2 or 4 1 2 480 across 32 enclosures 192 TB (64 TB per X-Blade)Figure 2. Scalability specificationsfor Dell/EMC NS storageDell/EMC NS-960 2 to 8 1 or 2 2 960 across 64 enclosures 896 TB (128 TB per X-Blade)Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 35

FeaturesectionSolutions for the virtual eraExploringinformation lifecycle managementefficienciesUsing multi-protocol, multi-tierDell/EMC storage systemstogether with advancedfeatures in Oracle Database 11gcan help increase efficiencywhile maintaining highperformanceoperationsand access. In this Webcast,Annette Cormier and EricCannell examine ways toalign infrastructure costswith the value of informationthroughout its life cycle.dellenterprise.com/goto/emcoracleBoth block- and file-performance throughputcan also be easily scaled. For block scaling,Dell/EMC NS storage arrays come with flexible,multi-protocol I/O ports that utilize the Dell/EMCUltraFlex architecture. With UltraFlex, I/O portscan support protocols such as iSCSI and FibreChannel as well as expected technologies suchas Fibre Channel over Ethernet (FCoE), and canbe upgraded or expanded online for increasedthroughput without application disruption—forexample, when moving from 4 Gbps to 8 GbpsFibre Channel or from 1 Gbps to 10 Gbps iSCSI.File-system (NAS) access throughput can beincreased as well, by adding additional X-Blades(data movers) to the storage array.Optimizing for efficiency and extendingcost-effective storageIn addition to providing scalable, highperformanceunified storage for sharing filesystem-leveland block-level storage, Dell/EMCNS unified storage systems are designed todeliver additional efficiencies such as reducedcapacity requirements, minimal power usage,and increased performance. Key features includethe following:• Virtual provisioning: To help ensure that endusers do not run out of capacity, organizationsoften overprovision storage, which can1 TBTraditional file dataresult in substantial underutilization of diskresources. Virtual provisioning, available onall Dell/EMC NS unified storage systems,enables storage capacity to be allocated simplyand automatically on an as-needed basisas opposed to all at once, which can helpdramatically increase disk utilization and lead toreduced power consumption costs.• Data deduplication and compression: Duplicateand infrequently used files can consume extracapacity and network bandwidth during bothstorage operations and backup and recoveryoperations over a local area network (LAN) orwide area network (WAN). Data deduplicationand compression features available withDell/EMC NS storage can reduce data capacityrequirements by up to 50 percent, helping tofurther reduce costs and optimize performance(see Figure 3).• Efficient drives: Low-power SATA II drivesand power-efficient EFDs help minimizeenergy costs in both high-capacity and highperformancesituations. Low-power SATA IIdrives require up to 32 percent less energythan traditional 1 TB, 7,200 rpm SATA drives forefficient, high-capacity storage; EFDs require upto 98 percent less energy per I/O per second(IOPS) than Fibre Channel drives for efficient,extremely high-performance storage.• Tiered storage: By enabling organizations tomix drive types within a single scalable storagesystem, Dell/EMC NS storage arrays helpsimplify the deployment and management oftiered storage and archiving infrastructures.Figure 3. Reduced capacityrequirements using datadeduplication and compressionfeatures available in Dell/EMCNS storage100 GBActivedata100 GBActivedata900 GBInactive or aged data orspecifically targeted data400 GBDeduplicated andcompressed data500 GBDeduplication-enabledfile systemUp to 500 GBof savingsDeploying built-in capabilitiesand delivering high availabilityThe Dell/EMC NS family of unified storage systemsincorporates an extensive range of built-infeatures designed to streamline management,protect data, and ensure availability. For example,the Celerra ® Manager software included with allDell/EMC NS storage systems is a Web browser–based tool that enables administrators to easilydeploy, configure, and manage file-system andapplication-based storage.The Celerra Provisioning Wizard, an integralpart of Celerra Manager, enables storage to be36 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Figure 4. Automatedreplication and recoveryavailable with CelerraManagerdeployed in just a few clicks. The AutomatedVolume Management feature in Celerra Managerenables easy provisioning and optimization offile systems based on workload; storage can beautomatically provisioned and tiered across disktypes based on workload type and policies.The Celerra File-Level Retention (FLR)feature, available in FLR-Enterprise (FLR-E) andFLR-Compliance (FLR-C) versions, supports simplemanagement of compliance and regulatorypolicies. For example, FLR is designed to protectfiles from deletion or modification until a specifiedretention date, enabling organizations to create apermanent, unalterable set of files and directoriesand help ensure data integrity. FLR also enableswrite-once, read-many (WORM) functionality onfile systems for information governance, and meetsthe requirement of U.S. Securities and ExchangeCommission (SEC) Rule 17a-4(f) regarding electronicstorage of broker-deal records. 1 This requirementmeans that an FLR-C file system that has files in thelocked state cannot be deleted.Celerra Manager also offers robust integrationwith third-party backup, antivirus, and quotamanagement tools. For example, the Celerra EventPublishing Agent (CEPA) provides integration withleading quota management software (such asNorthern Parklife’s Northern Storage Suite) thatenables event notification following Celerra fileand directory actions.Comprehensive integrationwith VMware virtualizationDell/EMC NS unified storage systems supportcomprehensive integration with the VMwarevCenter platform for simplified management andprotection of VMware virtualized environments.Using the VMware platform enables ITadministrators to choose among the NFS, iSCSI,Fibre Channel, and FCoE storage protocols,and Dell/EMC NS unified storage enables themto support all of these within a single storageinfrastructure. IT departments would typicallybegin a virtualization project with the protocolthat most comfortably matches their skill sets—forexample, companies that have Fibre ChannelSAN infrastructures might carry out their initialdeployments on Fibre Channel, while those thatdo not have Fibre Channel experience might optfor IP networking through either NFS or iSCSI.The Dell/EMC NS unified storage infrastructurewould enable these organizations to deploywhichever of these protocols they prefer. UsingVMware Storage vMotion technology or arrayEnablingnext-generationtiered storageBob Laliberte, senior analyst atthe Enterprise Strategy Group,discusses how the EMC ® FullyAutomated Storage Tiering(FAST) technology availablein some Dell/EMC arraysenables highly automated,dynamic, cost-efficient useof storage resources.bitpipe.com/detail/RES/1272658016_706.html1 For more information on SEC Rule 17a-4(f), visit sec.gov/rules/interp/34-47806.htm.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 37

FeaturesectionSolutions for the virtual era5Dell/EMCclicksNS storage arrays help organizations to cost-effectivelymeet demanding service-level agreements and simplifymanagement with easy-to-use unified storage that supportsCIFS, NFS, iSCSI, and Fibre Channel protocols. The Dell/EMC NSfamily is not only easy to configure and use, but also enablesorganizations to provision file shares in less than five mouse clicks.Annette Cormier isa solutions marketingmanager for Dell/EMCstorage solutions with20 years of experience indeveloping and bringingto market enterprisestorage, networkmanagement, andsecurity products.Eric Cannell is a productmarketing managerfor Dell/EMC storagesystems with many yearsof experience developingenterprise technologyproducts.Brad Bunce is thetechnical marketingdirector for EMCmidrange storagesolutions. He has17 years of IT experiencein technical presales,IT management,and marketing.Learn moreDell/EMC storage:dell.com/emcDell/EMC eventsand Webcasts:dellenterprise.com/goto/emcfunctionality enables administrators to move virtualmachine deployments between protocols. CelerraManager enables end-to-end virtual-to-physicalmapping of virtual machines to storage resourcesin a Web browser–based tool that helps streamlinemanagement of virtualized environments.The VMware Site Recovery Manager AutomatedFailback feature enables end-to-end disasterrecovery of VMware virtualized environmentsthrough a VMware vCenter plug-in, and theVMware View Storage Integration feature enablesrapid provisioning of virtual desktops through anadditional vCenter plug-in. VMware-aware datadeduplication and compression features arealso available with Dell/EMC NS storage arraysand provide comprehensive integration withVMware View, enabling data deduplication andcompression of VMware virtualized storage.Advanced data protectionand high availabilityTo help maximize file-system and applicationuptime, Dell/EMC NS unified storage systemsare designed for enterprise-level availability,and include advanced reliability capabilitiessuch as RAID technology for drive-level dataprotection, redundant storage processors forcontinued operation in the event of a failure,and active/passive failover. Dell/EMC NS storagearrays also support the Data Access in Real Time(DART) OS, which offers dynamic failover to ahot standby X-Blade (data mover) to help ensurethat data remains highly available and protected.Dell/EMC NS unified storage systems alsocome with robust data protection software,including automated snapshot, replication, andrecovery capabilities. Celerra SnapSure softwarecan create application-consistent point-in-timesnapshots, including writable snapshots for testand development (see Figure 4). Celerra SnapSurealso supports built-in automated recovery andutilizes pointer-based copy technology to helpminimize disk space and enhance performance.Celerra Replicator software supportsautomated, multisite replication. Replicator createspoint-in-time read-only and read/write copies ofproduction file systems or logical units (LUNs) onlocal or remote storage systems, and uses standardIP-based networks for maintaining consistentreplicas between sites. Replicator is designed forsimple administration using easy-to-define businesspolicies, and includes features such as applicationlevelrecovery point objectives and customizablebandwidth usage schedules. Bandwidth throttling isalso supported for situations when it is important toshare network bandwidth with other applications.Acting as a synchronizing or orchestration tool,Replication Manager resides at the host andleverages integration with Microsoft and Oraclesoftware to provide application-consistent replicasof Microsoft Exchange, Microsoft SQL Server,and Oracle Database data to help ensure thatsnapshots and replicas remain consistent forrecovery operations.Benefiting from consolidationon unified storageConsolidating block- and file-level storage intoa unified storage system has many potentialbenefits. The Dell/EMC NS family of unifiedstorage combines scalable, flexible, highperformingstorage with advanced efficiency, dataprotection, and management features that canhelp reduce total cost of ownership, enhanceavailability, and streamline the manageability ofenterprise storage deployments.38 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Solutions for the virtual eraFeaturesectionRacking updata center efficiencyBy Joyce RuffDesigned to meet the needs of high-density data centerenvironments, the Dell PowerEdge 4820 rack enclosurecan hold and protect a wide range of IT equipment whilesupporting enhanced power management, efficient cooling,and simplified component storage and mounting.When organizations deploy or expand their datacenters, they often struggle with finding the rightbalance between server capacity and powermanagement. Many choices depend on the spaceavailable in the room being used for equipment storage, and ITmanagers must also account for weight, power, cooling, and datamanagement requirements to help ensure that the data centeris both functional and efficient. One key decision is the selectionof a rack infrastructure: racks have become far more than simplecabinets, and the inclusion of features that help organizations store,power, cool, manage, and secure their hardware makes them a keypart of the overall data center design.Dell PowerEdge racks include an array of features to address thecritical power, cooling, and cabling issues that many organizationsface in their data centers. The Dell PowerEdge 4820 rack enclosureis the latest addition to the PowerEdge rack family, which nowincludes three models to help meet different needs: the24U PowerEdge 2420 rack, the 42U PowerEdge 4220 rack,and the 48U PowerEdge 4820 rack. Each rack is 600 mm wideand 1,070 mm deep to fit within a two-tile floor plan. Designedto hold and protect server, storage, and networking equipment,the PowerEdge 4820 provides a high-density option fororganizations with high-performance computing centers and similarenvironments that need to support additional servers while stillmaintaining the existing footprint.The Dell PowerEdge rack familyincludes three models to help meeta variety of data center needsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 39

FeaturesectionSolutions for the virtual eraThe 10-second challengeThe Dell ReadyRails mounting interface is designed for easyinstallation—with no tools required. In this video, see justhow quickly this design can be fitted onto a Dell rack.youtube.com/watch?v=7GCPOKSNuC4Designing for flexibility, strength,and securityBuilt with adjustable vertical mounting rails within therack, Dell PowerEdge rack enclosures are designedto accommodate multiple types of IT equipment.Because they adhere to the EIA-310-E standardfor rack mounting of electronics, the racks canfit not only Dell PowerEdge servers, but also DellEqualLogic , Dell PowerVault , and Dell/EMC storage;Dell PowerConnect switches along with othernetworking equipment such as routers, switches,and hubs; and even telephony equipment.The PowerEdge 4820 has a static load ratingof 2,500 pounds (1.13 metric tons), enabling it tohold a typical complement of equipment withoutthe need for special infrastructure. To help increaserigidity, IT departments can take advantage ofaccessories such as side stabilizer bars to securethe rack to the floor, as well as interconnect kitsto bolt adjacent racks to each other. Rotating rearcasters and accessible leveling feet on Dell racksenable easy maneuvering and positioning. Andbecause they incorporate lockable doors at thefront and rear of the rack as well as lockable sidepanels that use the same key, these racks can beused in almost any environment—including datacenters, remote offices, wiring closets, and evenfactory floors.Incorporating key accessoriesAlthough rack enclosures provide the necessarycapacity for holding key data center components,it takes more than just a rack to create an efficientdata center. To help organizations implement bestpractices for efficiency, Dell racks support a rangeof complementary accessories to enhance powermanagement, airflow, and component storageand mounting.To enable IT staff to mount Dell half-heightand full-height managed power distribution units(PDUs) without cables impeding airflow andequipment access, the Dell PowerEdge 4820 rackprovides a large distance between the back panelof the server and the PDU outlets. This rack alsosupports a wide range of options for PDU typesand form factors, and includes a tray for mountingPDUs at the rear of the rack as well as support forinstalling them in the U-spaces. The PDU tray hasbeen enhanced with additional mounting locationsfor PDUs and cable management accessories.Rack fan kits and blanking panels can helpensure proper airflow in the rack and enhancecooling efficiency. Plastic blanking panels in 1Uand 2U sizes are now available in addition to steelblanking panels in 1U, 2U, 3U, and 6U sizes. Theplastic blanking panels have a tool-less snap-indesign that enables quick, easy installation andremoval in the unused U-spaces in a square-holerack. The screw-in design of the steel panelsprovides support for a wider range of racks,including threaded and unthreaded round-holeracks as well as square-hole racks.IT staff can quickly and easily mount latestgenerationPowerEdge servers in PowerEdge rackenclosures using the Dell ReadyRails tool-lessmounting interface, which includes spring-loadedlatches designed to engage automatically. Thisefficient design makes the release latch visibleand accessible from the front of the rack, withoutrequiring special tools or empty U-spaces above orbelow to disengage them.These rail kits are available in sliding or staticstyles to accommodate the needs of differentserver specifications and the environmentas a whole. Sliding rails allow the system tobe fully extended out of the rack for service,40 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

and additionally support an optional cablemanagement arm (CMA) that attaches without theuse of tools and provides a guide for componentcable routing to the rear of the rack. Becausestatic rails are less complex than sliding rails anddo not need CMA support, they offer a smallerfootprint. Both types of ReadyRails mounting kitscan be installed in EIA-310-E-compliant four-postsquare-hole and unthreaded round-hole racks; thestatic rail kits can also be mounted into four-postthreaded-hole racks and two-post racks.For components that do not come with railsfor rack installation, IT staff can take advantage ofthe 1U fixed equipment shelf, which is designed fortool-less installation into square-hole or unthreadedround-hole racks using the ReadyRails mountinginterface. The rail design for the shelf also supportstooled installation in four-post and two-postthreaded-hole racks for added versatility. This steelshelf is designed to hold up to 200 pounds ofweight, and comes with a pair of hook-and-loopstraps to secure cables to the shelf.Enabling efficient,high-density environmentsServer racks are an important part of effective datacenter design. In conjunction with key accessories,the Dell PowerEdge 4820 rack enclosure providesa flexible way for organizations to create efficient,high-density environments without increasingfloor-space requirements.Learn moreJoyce Ruff is a product marketing managerfor the Data Center Infrastructure teamwithin the Dell Enterprise Product Group.Dell racks and accessories:dell.com/poweredge/rackReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.BEST PrACTICES fOr rEDUCINgNETwOrkINg COmPLExITyLearn More on Page 91GEt thE MoStBanDwIDth out ofYour Data CEntEr.nEw Intel® EthernetX520 10GbE Dual PortMezzanine CardThe new Intel® Ethernet server and blade adapters feature advanced I/O virtualization, native iSCSI acceleration, and iSCSI Boottechnologies for customers who need high levels of performance for their LAN and storage connections.Intel® Virtualization Technology for Connectivity reduces I/O bottlenecks in a virtualized server and improves performance, flexibility and virtualmachine scalability. Optimized for virtualization, Intel’s native iSCSI acceleration solution delivers iSCSI performance using customer-preferred,native iSCSI initiators and intelligent stateless offloads, making it the easiest and most cost-effective way to connect servers to iSCSI SANs.Intel® Ethernet Adapters on Dell PowerEdge servers—visit www.IntelEthernet-DellPS.com to learn more.nEw Intel® EthernetX520-Da2 10GbE Dual PortServer adapterIntel® 10 Gigabit Xf SrServer adapterIntel® 10 Gigabit atServer adapterIntel® Gigabit EtDual Port Server adapterIntel® Gigabit EtQuad Port Server adapter2010, Intel Corporation. All rights reserved. Intel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries.©Intel® Gigabit EtQuad Port Mezzanine Cardfor M-Series Blade Servers

Customer perspectiveTsumura & CompanySolving IT problemswith server consolidationBy consolidating its infrastructure on virtualized Dell PowerEdge servers, Japanese herbal medicine leader Tsumura & Company haslaid the groundwork for a dramatic reduction in ongoing servercosts while ensuring that legacy applications can continue runningyears into the future.With its motto of “The Bestof Nature and Science,”Tsumura & Companyaspires to contribute tohuman health and well-being by fusingWestern medicine with a scientificallygrounded understanding of kampo, Japan’straditional herbal medicines. In addition toits 129 kampo products, Tsumura provides136 varieties of healing medicines and makessignificant on-site contributions at a widevariety of medical facilities.Tsumura has long been committedto advanced technology initiatives, but itsrelentless focus on IT had led to a problem: anexplosion in the number of in-house systems,particularly servers. “Even as we continue tointroduce more and more servers,” says HideoSato, head of the Tsumura IT department, “oneby one our existing servers are approachingthe end of their lives, and the effort and costneeded to update our in-house systemskeeps on growing. In this environment I findmyself constantly wishing I could bite downand put a stop to things.”Reducing fixed operating costsWith an eye toward reducing costs, Tsumurawent in search of a game-changingsolution—and found one in virtualizationbasedserver consolidation. Tsumura had firstimplemented VMware ® server virtualizationaround 2005, but at that time, the hardwareand software had not yet matured enough tosupport true consolidation.42 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

InnovativesavingsAdvanced IT initiativeshave been a focus atTsumura & Companyfor years. Now, servervirtualization andconsolidation on DellPowerEdge serversare paving the wayfor future growth.137:12 68.2% 224 tonsFollowing a DellVirtualization Assessment,Tsumura found that itcould consolidate 137 ofits Web and applicationservers down to just 12virtualized Dell servers.The company’s preliminaryestimates indicatethat the consolidationproject could reduce itsannualized server costs by68.2 percent.In addition to the costadvantages, Tsumurareports that the projecthas contributed toenvironmentally friendlyIT practices—reducing itspower consumption by anamount equivalent to 224tons of CO 2 emissions.Big changes were in store in 2008.Dell and Mitsubishi Research Institute DCS,the two outsourcing partners in charge ofTsumura’s system operations, proposedthat the company deploy the latest versionof the VMware platform. Two featuresthat particularly impressed Sato werethe VMware vMotion and VMware HighAvailability technologies. “Previously, ourprimary server machines had never gottenbeyond a redundancy of about 50 percent,”he says. “If the server consolidation we getfrom the latest version of VMware enablesus to reach an effective redundancy of near100 percent, then I would say that wasdefinitely something worth trying!”The targets for this round of consolidationwere 137 Web and application servers. Afteran initial Dell Virtualization Assessmentservice, Tsumura saw that it couldconcentrate its resources on just 12 DellPowerEdge servers. Preliminary estimatessuggested that the consolidation couldreduce the combined annualized costs forservers and server farms by 68.2 percent—ahuge savings for the company.The Dell team was able to get the newenvironment up and running quickly. “AtTsumura, once we made the decision, wewanted to execute it immediately,” saysTakuo Yamaguchi, leader of the informationmanagement group within Tsumura’s ITdepartment. “Dell was there to meet our“Dell was there to meetour needs—in lessthan six months, anincredibly short interval,they had implementedour server consolidationproject.”—Takuo YamaguchiLeader of the information managementgroup at Tsumura & CompanySeptember 2009needs—in less than six months, an incrediblyshort interval, they had implemented ourserver consolidation project.”In addition to helping significantlyreduce costs, virtualization is also helpingpreserve system continuity well into thefuture. “Previously, we had been pretty muchredoing our whole system every five yearsor so,” says Sato. “The truth is, once youreplace the hardware, you wind up havingto upgrade to newer versions of the OS andthe middleware, and in order to maintaincompatibility you then wind up with nochoice but to redo the entire system.” Thevirtualized infrastructure enables Tsumurato maintain compatibility across all of itsoperating environments—a reassurance thatmeans a lot to the company.Consolidation has also brought majoradvantages in another area: environmentallyfriendly IT. “By significantly concentratingour servers and other hardware resources,we have reduced our electric powerconsumption by an amount equivalent toaround 224 tons of CO 2 emissions,” says Sato.Moving toward database serverconsolidation“Above all else,” says Sato, “the point onwhich I really want to commend Dell is thedriving force they provided for this project.In collaboration with the technical staffat Mitsubishi Research Institute DCS, theyproceeded in an extremely short timefrom proposal to design, architecture, andfull transition.”Yamaguchi agrees. “I’m looking forwardto two or three years from now, when we’llbe able to assess how the cost reductionswe achieved through this round of serverconsolidation took shape and stimulated ourmanagement and our business.”As a next step, Tsumura plans toconsolidate its database servers. Incontrast to the relatively compact Weband application servers, these servers arelarge-scale systems that demand high levelsof scalability, and several hurdles remainto be cleared. But with the experienceand knowledge gained thus far, the firm isoptimistic that it can meet the challenge.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 43

Customer perspectiveState of DelawareTaming the e-mail beastwith bladesCentralizing its e-mail systems on Dell PowerEdge bladeservers and Dell EqualLogic storage helps the state of Delawarereduce total cost of ownership, streamline management, and cutdowntime from days to minutes.The second-smallest U.S. state after Rhode Island, Delaware is a 30-by-96-milewedge of land on the Delmarva Peninsula. The state government maintains acentral data center in the capital of Dover and a disaster recovery facility in NewCastle, and its e-mail system processes more than a million messages a day.A decade ago, the state began consolidating its systems as part of a move to theMicrosoft ® Exchange and Active Directory ® platforms. The government initially chose HP bladeservers, but recently switched to Dell PowerEdge M-Series blade servers with Intel ® Xeon ®processors. “When we migrated to Microsoft Exchange Server 2007, Dell presented us anCutting costsThe state ofDelaware is alreadyseeing results fromits consolidationon Dell PowerEdgeblades and DellEqualLogic storage.16:4The state initially required16 clusters to run MicrosoftExchange; now 4 clustersof PowerEdge M610 bladeservers get the job done.50%The easy-to-useEqualLogic arrays have cutstorage management timeby 50 percent.outstanding value for a consolidation solution,”says Douglas Lilly, lead telecomm technologistfor Delaware’s Department of Technology andInformation. “When we first started out withMicrosoft Exchange, we had 16 clusters, whichwe reduced to 12 clusters of HP server blades.Now we’ve reduced those to 4 clusters ofDell PowerEdge M610 blade servers.”The Dell blades have brought a varietyof benefits, including reducing downtimeeach year from days to minutes, reducingrack space requirements by 66 percent,simplifying remote deployment andmanagement, lowering energy costs,and significantly increasing capacity. Lillyestimates that matching the current capacityon the previous systems would raise totalcost of ownership by about 20 percent.Dell EqualLogic PS5000E storagearrays help meet the state’s large e-mailvolume and retention requirementswhile streamlining storage management.“Compared to other storage, the ease ofadministration is fabulous,” he says. “WithEqualLogic, it takes 50 percent less time toadminister LUNs and set up the servers withthe storage.” He estimates that the simplifiedconnectivity saves US$1,000 per server port.Lilly also praises Dell support services.“We really like Dell services and products,which is why we are running our entirestate Active Directory, e-mail infrastructure,Microsoft Office Communications Serverdeployment, and Microsoft SQL Server ®databases on Dell PowerEdge blade serverswith Intel processors,” he concludes.44 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Systems managementDell KACE appliances:Simplifying Windows 7 migrationsBy Sendhil JayachandranFor midsize organizations workingwith a diverse range of hardware andlimited resources, making the moveto the Microsoft ® Windows ® 7 OS canbe a daunting challenge. The innovativeDell KACE family of appliancesprovides a simplified, automated,cost-effective way to design and carryout an end-to-end migration.Enterprise IT departments are quickly embracing the MicrosoftWindows 7 OS: in a recent survey, 87 percent of organizationssaid that they plan to deploy Windows 7, and almost halfdo not plan to wait for Service Pack 1 before rolling it out. 1Because Microsoft does not support in-place upgrades of Windows XPto Windows 7, however, the migration can pose a daunting challengefor organizations that bypassed the Windows Vista ® OS and continuedrunning Windows XP, but still need a means to execute clean installationsof Windows 7 while retaining critical data.The migration can pose an even larger challenge for midsizeorganizations, which may need to carry it out with limited staff andfinancial resources. These organizations may also be highly flexible aboutthe types of hardware and operating systems allowed on their networks,requiring IT administrators to account for a diversity of systems andOS platforms during the rollout. When planning a Windows 7 migrationstrategy, these organizations should follow best practices that align withtheir specific environment and use tools that can help them to optimizethe migration process. Designed specifically with the needs of midsizeorganizations in mind, the Dell KACE family of appliances offers a way foradministrators to create a simplified, automated, cost-effective approachto Windows 7 migration.1 “Windows 7 Adoption: A Survey of Technology Professionals,” by Dimensional Research, January 2010,available at kace.com/resources/windows-7-adoption-survey-2010.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 47

Systems managementFigure 1. Scripted installation detailsin the Dell KACE managementinterfaceSuccess storiesDell KACE appliances haveenabled a wide range oforganizations to dramaticallysimplify ongoing systemsmanagement. In this series ofvideos, customers explainhow these appliances havechanged the way they runtheir IT operations.dell.com/kacecustomersAssessing hardware andsoftware compatibilityOne of the first steps to executing a successfulWindows 7 migration is to assess the hardware andsoftware in the environment to determine readinessand potential upgrade needs. IT departments shouldstart by determining the mix of systems runningWindows, Linux ® , or other operating systems. Therelease of Windows 7 presents an opportunity toincrease standardization: administrators mightdecide that they want to remove an OS from theirnetwork, or at a minimum ensure that all usersare on the same version of each OS. Taking thetime to standardize platforms can be a worthwhileinvestment for midsize organizations, resulting insimplified systems management.In the same survey referenced earlier,software compatibility was identified as thesingle greatest concern organizations hadregarding Windows 7. Automated tools doexist that can assist with verifying Windows 7software compatibility, and these tools workwith centralized clearinghouse databases ofalready-tested applications. Just as they canstandardize their OS platforms as part of themigration process, midsize organizations canalso take this opportunity to consolidate theirsoftware applications. Critical applications thatcan’t be consolidated and aren’t compatible withWindows 7 can be run using Windows XP Mode,a feature available in the Windows 7 Professionaland Ultimate editions that uses an updated versionof Windows Virtual PC to maintain a virtualizedWindows XP environment within Windows 7. Theprocess is hidden from end users: applicationsappear as if they are running on Windows 7, whilein fact they are running in the background on theWindows XP virtual machine. Because the useof Windows Virtual PC requires special hardwarecapabilities, administrators should be sure toinventory their systems to ensure that they arecompatible with Windows XP Mode.Optimizing OS deploymentMidsize organizations need a way to streamlinethe Windows 7 deployment process as much aspossible to minimize manual work. An optimaldeployment should incorporate gold masterautomation, hardware-independent deployment,user state migration, and cross-platform support.To deploy Windows 7 throughout anorganization, administrators typically build a goldmaster reference system from which to capture adisk image—a process that can consume precioustime and resources if executed manually.Automating the process can help to both simplifyimage captures and increase reliability, avoidingthe need to rely on manually inserting theappropriate installation media into the referencesystem. For midsize organizations that may nothave the time, personnel, or capital to dedicatestaff toward building and maintaining thesereference systems, gold master automationcan be a necessity.An optimized Windows 7 migration shouldalso allow for a driver repository of knownhardware, to enable the creation of a hardwareindependentinstallation process that canautomatically customize itself at the point ofinstallation for each hardware class. For midsizeorganizations that need to accommodate awide range of systems, hardware-independentinstallations become particularly useful overtime—enabling repeated use of the sameinstallation as new hardware is purchased. Afterthe initial setup, administrators should need toconcern themselves only with ensuring that newdrivers are contained in or added to the databaseas needed for each deployment.User state migration allows user-specific filesand settings to be deployed along with the OSand applications, helping reduce the risk of losingcritical information and helping minimize enduserdowntime. This step applies to both cleaninstallations on existing systems and migrations tonew systems; a successful migration means thatwhen users turn on their new system or OS for thefirst time, the information they want to retain fromtheir old system or OS is already there. For midsizeorganizations that may back up end-user datainconsistently, it is particularly important to migrateuser states to help ensure that critical data is notlost. By collecting user states on a regular basisafter a Windows 7 migration, administrators canpotentially refresh systems at any time, rather thanjust during the migration project. Users in such48 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Dell KACE systems management appliances fulfill all your PC andserver management needs, from initial computer deployment toongoing management and retirement.Easy-to-Use• Intuitive tabbed web-based console• Integrate with LDAP, Active Directory and email• One-click updates simplify upgrades and patchesComprehensive• Full PC and server lifecycle management• Dell KACE supports Windows, Mac and Linux operating systems• Deploy in one day• No hardware or software prerequisites• No professional services requiredSystem requirementsVMware ESX 4 or later / vSphere• Intel-VT or AMD-V 64 bit processors• 4 GB RAM• 250 GB minimum disk spaceVMware ESX 3 or later / viClient• Intel-VT or AMD-V 64 bit platforms• 4 GB RAM• 250 GB minimum disk spaceVMware Player 2.5.3 or later(Trial Dell KACE appliances only. Not supported byproduction Dell KACE virtual appliances.)• Windows-based system with Intel-VT or AMD-Vcapable processor(s)• Windows XP or later, either 32- or 64-bit• 2 GB RAM• 7 GB disk space minimum (additional space required to testbackup features; approximately 50 GB to test the patching feature)*VMware, ESX, Player, vSphere and viClient are registered trademarks ofVMware, Inc. in the United States and/or other jurisdictions.

Dell KACE appliances are designed for exceptional performance,reliability, ease of use, and scalabilityan environment may choose to have theirsystems seamlessly rebuilt when a problemcannot be solved. The same holds true aftera crash, helping administrators to quicklyprovision a new system with little or no lossof user data.Finally, administrators in midsizeorganizations should utilize a tool thatprovides cross-platform support forWindows, Linux, and other operatingsystems. In heterogeneous environments,using tools that can manage only a singleplatform would require deploying separatetools for each supported OS—which canresult in increased infrastructure andtraining costs that the organization may notbe able to afford.Taking an appliance-basedapproach to migrationThe Dell KACE family of physical and virtualappliances offers an innovative approach toWindows 7 migration specifically designedfor midsize organizations—providing an easyto-use,comprehensive, and cost-effectiveway to carry out Windows 7 migrations.KACE physical appliances are based on 1UDell PowerEdge rack servers, while KACEvirtual appliances can run in the VMware ®ESX and ESXi virtualization platforms.Both physical and virtual appliances aredesigned to seamlessly integrate OS,database, file server, Web server, andsystems management software to providea true plug-and-play solution that supportsWindows, Linux, and Apple Mac OS Xoperating systems (see Figure 1).Windows 7 migrations can entailmuch more than just the OS deploymentitself, including inventory readinessanalysis, user state migration, softwareand patch distribution, and ongoing assetmanagement. By providing comprehensivesupport for the process as a whole, theseappliances can help eliminate the need formidsize organizations to task IT personnelwith tactical work, allowing them to focusinstead on strategic initiatives.Midsize organizations also may not havethe luxury of investing in consulting servicesand extended training. KACE appliancescan be plugged into an existing networkand immediately begin functioning, helpingsupport rapid, simplified deployments.The appliance-based architecture helpseliminate many of the costs of traditionalsystems management software packages—there are no professional services fees toincur, no additional hardware or softwarerequirements, no custom integration costs,and only minimal training requirements. Atthe same time, KACE appliances can provideexceptional performance, reliability, easeof use, and scalability in a purpose-builtappliance that is pre-tuned, hardened,and self-monitoring.Enabling simplified, automated,cost-effective migrationsSteve Brasen, a senior analyst with ITresearch firm Enterprise ManagementAssociates, offers this perspective on theWindows 7 migration challenge: “In thehistory of Windows, we’ve never seen thekind of rapid adoption of a new OS versionas we have with Windows 7. The moveto Windows 7, however, is not a processthat can be effectively performed throughmanual processes. Automated migrationsolutions such as Dell KACE deploymentappliances are essential for ensuringconsistency, reliability, and cost-effectivenessin a migration strategy.”This perspective can resonate stronglywith midsize organizations in particular,which are often contending with diversehardware, limited IT budgets, and personnelconstraints that make ease of use andcost-effectiveness essential. Throughtheir innovative approach to Windows 7migration, Dell KACE appliances can helpthese organizations simplify, automate, andoptimize the move to Windows 7 as part ofan end-to-end migration strategy.Learn moreSendhil Jayachandran is a senior productmanager at Dell KACE, where he isresponsible for product managementactivities related to the Dell KACE K2000deployment appliance.Dell KACE:dell.com/kaceDell KACE free trial:dell.com/kacetrialDell and Windows 7:dell.com/windows7enterpriseDell KACE and Windows 7 migration:dell.com/kacewindows7Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 49

Storage optimizationEnhancing SANself-management withDell EqualLogic storageBy Tony Ansley and Mansour KaramDell EqualLogic arrays offer built-in intelligence tohelp simplify configuration—but operators may still missimportant settings. Together, Arista and Dell have createda storage architecture that can automate configurationthrough switches running the Arista ® Extensible OperatingSystem (EOS ® ) platform.Ever-increasing growth in the volume ofstored data, coupled with applicationsthat demand large computational power,are driving the need for scalable, highcapacitystorage area networks (SANs). At the sametime, a constant push to simplify IT infrastructureand reduce operating costs is driving a move tovirtualize operations in the storage segment.These factors put a premium on interconnectbandwidth and manageability. Scaling to hundredsof terabytes or petabytes of storage capacityrequires switches that can cost-effectively providehigh bandwidth and low latency. To help simplify ITmanagement and minimize operating costs, tightintegration is necessary between the managementcapabilities of storage platforms and those of theswitches. Successful operation requires a highdegree of automation and integration into anorganization’s IT environment.The combination of Dell EqualLogic PS SeriesInternet SCSI (iSCSI) SAN arrays and the Arista7000 Family of Gigabit Ethernet (GbE) and10 Gigabit Ethernet (10GbE) switches helps meetthese requirements. Now available with 10GbEcontrollers, the EqualLogic PS6000 series ofarrays can be deployed in conjunction with Aristaswitches to provide a cost-effective 10GbE-basedcloud storage platform that is well suited tobandwidth-intensive software such as database,analytics, search, high-performance computing(HPC), and media applications.Simplifying management throughautomatic configurationDell EqualLogic PS Series iSCSI SAN arrays offer highlevels of built-in intelligence and self-management.They are designed to automatically load balancedata across network connections, disk drives, andarrays to help optimize performance, and canautomate storage configuration, management,pooling, storage tiers, and data distribution to helpsave time and increase productivity.Switches from Arista Networks offer acustomizable, extensible switch OS thatprovides a platform for automation in supportof EqualLogic storage—and combining the twoenhances the self-management capabilities ofboth platforms. The Arista switch software hasdistinct properties that make this high level ofself-management possible.50 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Many of today’s network switches presenta fundamental design limitation: a single defectanywhere in the OS can potentially expose theentire system to disruption. At the same time, withno isolation between multiple tasks, it is difficultto add functionality. In fact, making any significantchanges to a code base that is measured inmillions of lines of code is likely to reduce productstability and reliability.The fragile nature of this monolithic approachmakes it inherently difficult to extend thenetwork OS to implement new switch featuresor provide integration with other systems. TheArista Extensible Operating System (EOS) softwareavailable across the Arista 7000 Family of switchesis designed to avoid this limitation.Built on a Linux ® kernel, EOS can be easilyextended to implement new features suchas integration with storage-array-specificmanagement processes. Tight integration betweenthe EqualLogic management firmware and EOSallows Arista switches to automatically detectwhen an EqualLogic SAN is connected andoptimize the switch configuration accordingly.The process requires minimal manual intervention,which helps dramatically reduce the risk of humanerror and associated costs.Understanding the modularArista EOS architectureThe key to EOS extensibility is its state-sharingarchitecture (see Figure 1). This architectureconsists of multiple processes interacting with acentral shared-state repository called Sysdb (shortfor system database), a design that separates thenetworking state from the actual processing.Each switch function is in a separate addressspace, including command-line interface (CLI)sessions, hardware device drivers, and individualprotocols such as the Open Shortest PathFirst (OSPF) routing protocol, Spanning TreeProtocol (STP), and Link Aggregation ControlProtocol (LACP). Sysdb holds the stateinformation, while agents implement the logicand perform the processing. Sysdb delivers stateupdates from one agent to another, and stateinformation is recoverable from Sysdb whenneeded. Like a traditional database engine,Arista EOS:Open to extensionAdding functionality to network switches can be difficult forIT organizations or third parties. In the spirit of openness, theSysdb programming model and application programminginterface (API) set in the Arista Extensible Operating System(EOS) software is visible and available through a standardshell. This approach offers a number of advantages:The Sysdb API set is not a “walled garden” that includesonly a limited subset of features; all APIs that Arista softwaredevelopers use between address spaces within EOS areavailable to third-party developers and organizations usingArista switches.Developers can create third-party agents to customizeswitch behavior and automate common tasks such asdeployment, monitoring, maintenance, and upgrades withinspecific IT environments.EOS has Linux shell access for root-level administrators, andmakes a broad suite of Linux-based tools available (such astcpdump and fping).The extensible EOS architecture enables enhancedcapabilities with Dell EqualLogic PS Series storage arrays.Sysdb is designed to be reliable because itcontains no application code.All interprocess communication in EOS isimplemented as writes to Sysdb objects. Thesewrites propagate to the appropriate agents,triggering events in those agents. For example, ifa user-level application-specific integrated circuit(ASIC) driver detects a link-down state, it writesthis to Sysdb. An LED driver would then receive anupdate from Sysdb and adjust the LED status toreflect the link-down state.EOS helps provide reliable interprocesscommunication services while preserving theLinux heritage of openness and extensibility (seeReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 51

Storage optimizationoptimal support for that array. DCBX can be used toXML andSOAPLAGSNMPCLISysdbiSCSIDCBX/PFCautomatedconfigurationASICdriversVirtualswitchmanagementprioritize the handling of iSCSI traffic to help ensurethat packets are not dropped or delayed.Using Arista’s application programminginterface (API) set for Sysdb, administratorscan create their own separate agents that cancommunicate with, read from, and write to Sysdbto implement additional functionality. Theseextensions are designed to be safe and reliablebecause the switch state is protected within Sysdb,and faults within the extensions are contained.Simple agents can be implemented to extendautomation of the network configuration toinclude LACP groups or STP parameters.STPLinux kernelHardwareOSPFFigure 1. Arista EOS architecture consisting of multiple processes interactingwith the central Sysdb shared-state repositorythe “Arista EOS: Open to extension” sidebar in thisarticle). Protocol operation, management function,and device management—including hardwaredevice drivers—run in user address spaces ratherthan in the kernel. Keeping the vast majority ofprocessing out of the kernel helps to increasesystem stability and simplify extensibility. Addinga user process to a Linux-based platform can besimpler and safer than adding kernel-level code.EOS implements Link Layer Discovery Protocol(LLDP) and the Data Center Bridging CapabilityExchange (DCBX) protocol to help automatethe configuration of Data Center Bridging (DCB)parameters, including the upcoming Priority-BasedFlow Control (PFC) standard, which is expected toallow an end-to-end flow-control feature. DCBXis an extension of LLDP that allows storage arraysand switches to exchange capability parameters—coded in a type-length-value (TLV) format—andautomatically negotiate common PFC parameters.This feature enables a switch to recognizewhen it is connected to a Dell EqualLogic iSCSI SANarray and automatically configure the switch linkparameters (such as priority flow control) to provideDeploying Arista EOS withDell EqualLogic storageDell has developed best practices for infrastructureconfiguration that are designed to optimize DellEqualLogic PS Series iSCSI SAN deployments. Inparticular, these best practices revolve around howthe individual switches that make up that SAN areconnected.An Arista switch can configure an optimizedstorage infrastructure based on Dell best practicesand organizational requirements. The switchsoftware performs discovery and configuration in athree-step process:1. Peer switch discovery and configuration:Using DCBX/LLDP, the Arista fabricautomatically discovers peer switches (seeFigure 2). For typical SANs, an administratormust create a construct in the switch fabriccalled a link aggregation group (LAG), whichis a group of individual ports on the switchthat act as a single virtual link or port. Creatinga LAG for these ports can be automated byimplementing a simple EOS agent.2. Storage array discovery and configuration:Again using DCBX/LLDP, the Arista fabricautomatically discovers the ports on theEqualLogic arrays. EqualLogic arrays andArista switches both implement DCBX/LLDP,which allows the arrays to give the switchesinformation about their capabilities. Theconfiguration of appropriate settings for theseports is automated through DCBX, which52 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

[ Tough quesTion #3 ]How Does one of tHe worlD’slargest auDiting firms ProviDe80,000 secure connections?Today’s successful companies deliver critical applications in a reliable andsecure manner. SonicWALL security solutions provide granular control, proactiveprotection, and centralized management to meet these needs. With itsmulti-gigabit Reassembly-Free Deep Packet Inspection firewall, applicationintelligence and control, and SSL VPN leveraging a GRID network of4 million touch points, SonicWALL offers the flexibility to economically addresskey business challenges including mobility, cloud-based applications, andremote connectivity.learn more at sonicwall.com/80kstrongNETWORKSECURITYSECUREREMOTE ACCESSWEB AND E-MAILSECURITYBACKUP ANDRECOVERYPOLICY ANDMANAGEMENT© 2010 SonicWALL, Inc. SonicWALL and the SonicWALL logo are registered trademarks of SonicWALL, Inc.

Storage optimizationFigure 2. AristaEOS three-stepdiscovery and autoconfigurationprocessfor peer switches,storage arrays, andserversArista 7000Family switchesServer discoveryand configurationDell PowerEdge servers31Peer switch discoveryand configuration2Storage array discoveryand configurationDell EqualLogic PS Series iSCSI SAN arraysEnabling flexiblenetworking forvirtualizationVirtualization hasintroduced its ownnetworking challengesfor IT administrators.This video presents anextended overview of theArista vEOS platform, anEOS implementation thatintegrates with the VMwarevNetwork Distributed Switchfeature to help maintainconsistent operationalmodels and policies acrossthe environment.aristanetworks.com/media/system/video/AristavEOS_Detail.movautomatically configures the appropriatesettings for the PFC standard and gives priorityto iSCSI traffic.3. Server discovery and configuration: Ports withattached servers are discovered automatically.Based on the information obtained throughDCBX/LLDP, some settings for the serverlinks can be automatically changed byimplementing a simple EOS agent.Assessing performanceand scalability benefitsWith up to 48 non-blocking 10GbE ports in just1U of space, and 384 ports in an 11U form factor,the Arista 7000 Family of switches provides ahigh-performance, low-latency, space-efficientinterconnect for 10GbE Dell EqualLogic PSSeries iSCSI SAN arrays. Performance from theEqualLogic platform is optimized based on theswitching architecture that suits the needs of thespecific IT environment.The combined EqualLogic and Aristaarchitecture helps eliminate errors introducedduring traditional network configurations, potentiallysaving many hours of troubleshooting andremediation time for IT staff while also helping tominimize downtime for end users. The savings intime and labor because of simplified managementand automated optimization can help reduce thetotal cost of ownership of the SAN infrastructure.This architecture also enables simplifiedscaling to petabytes of capacity. For example, theEqualLogic PS6000 series of arrays is designed toscale up to 24 TB of storage per unit of rack space,and the Arista 7000 Family of switches is designedto scale up to 1 Tbps of capacity per unit of rackspace. This high density along with the automationfeatures in the EqualLogic arrays and Arista switchesmake it nearly as easy to deploy petabytes ofstorage as it is to deploy a single terabyte.Finally, EqualLogic arrays and Arista switchescan be combined in a power- and space-efficientway that helps reduce operating expenditures.Arista switches are designed to minimize powerconsumption, and the power supplies are designedfor highly efficient operation. IT organizations canachieve high densities by taking advantage of thescalable capacity of the EqualLogic PS6000 arraysand the high port counts of the Arista switches.Optimizing the architecturefor virtual environmentsThe combined Dell EqualLogic and Aristaarchitecture provides a unified managementframework for virtualized compute, networking,and storage resources. EqualLogic datamanagement is optimized for virtualization,and the Arista Virtualized Extensible OperatingSystem (vEOS) platform binds network policyconfigurations to storage and server links to helpkeep policies consistent.Arista vEOS is an implementation of EOS thatis designed to work with the VMware ® vNetworkDistributed Switch feature. Working together withthe VMware vSphere 4 platform, vEOS helpsto increase workload mobility. It preserves thenetworking state and policy as virtual machinesmove around the data center and into the cloud,providing visibility and consistency betweenphysical, virtual, and cloud networks. Using vEOS,administrators can treat virtual distributed switchesas an extension of the physical network and canmanage them using a CLI and Simple NetworkManagement Protocol (SNMP).Together, EqualLogic arrays and Arista 10GbEconnectivity are well suited to high-bandwidth,54 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

low-latency networking in today’s increasinglyvirtualized data centers. The architecture supportsefficient migration and replication to enhance theutilization, scalability, and flexibility of storage andserver resources.Minimizing fixed overheadfor infrastructure managementWith its ability to scale storage capacity andbandwidth requirements, and to help simplifymanagement across both physical and virtualizedenvironments, the joint Dell EqualLogic and Aristaplatform offers organizations an automated,self-managing storage infrastructure. EqualLogicPS Series iSCSI SAN arrays can automate muchof the day-to-day management of logical units(LUNs), volumes, and disks where data is physicallylocated. This built-in intelligence helps reducethe amount of knowledge and time requiredto manage performance and monitoring, whilehelping to ensure that the SAN is operatingoptimally for the application.Integration with Arista EOS adds value to thisdesign by further helping minimize the knowledgeand time required for setup and helping ensure thatthe configuration is resilient and can automaticallyadjust to changes in the infrastructure. Just asEqualLogic arrays are designed to reduce the timeand effort involved in storage management, anarchitecture that combines EqualLogic storage andArista switches can reduce overall infrastructuremanagement demands.Learn moreDell EqualLogic PS Series:dell.com/psseriesdell.com/equallogicArista EOS:aristanetworks.com/en/products/eosTony Ansley is a seniorstorage consultant on theiSCSI Solutions Marketingteam at Dell, and has over25 years of experience inthe computer industry.Mansour Karam isthe managing directorof worldwide businessdevelopment at AristaNetworks.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.SwitchN to Eaton.Power through high energy costs.Industry-leading backup power solutions from Eaton ®increase IT productivity and reduce costs. Withleading energy efficiency, simplified manageability,superior reliability, innovative designs and flexiblescalability, Eaton UPSs make it easy to be green.With Eaton’s highly efficient UPSs, you can slashenergy costs through reduced power consumptionand heat output. Eaton’s superior UPS technologyenables you to carry on efficiently — uninterrupted.www.eaton.com/powerqualityIt’s easy withEaton’s higheffi ciency UPSsolutions.Eaton is a trademark of Eaton Corporation. ©2010 EatonCorporation. All rights reserved. #11668_0610_TN

Storage optimizationSimplified monitoring ofDell EqualLogic storage withMicrosoft System CenterBy Viswanathan Balakrishnan, Stanley L. Stevens, Sekhar Duggirala, and Tom GeorgeThe Dell OpenManage Integration Suite for Microsoft ®System Center offers a simplified way to incorporate Dell hardware into the System Center management framework.As part of this software portfolio, the Dell EqualLogic Storage Management Pack Suite extends this functionality tohelp administrators efficiently monitor EqualLogic storage.Dell EqualLogicManagementPack in actionVisit this Dell TechCenterpage for a demonstration ofmonitoring Dell EqualLogicstorage with MicrosoftSystem Center OperationsManager.delltechcenter.com/page/system+center+operations+manager+demonstrationsAs enterprise data centers migratetoward increasingly consolidatedinfrastructures based on virtualizationand cloud networks, monitoring andmanaging all data center components—includingservers, storage, and switches—can becomeincreasingly complex, time-consuming, and costly.When these considerations are placed alongsidethe goals of maximizing productivity, providingflexible access to IT resources, and maintaining24/7 operations, the need for a comprehensive,integrated approach to systems managementbecomes clear.The Microsoft System Center suite includesa range of management tools that enable ITprofessionals to plan, deploy, manage, andoptimize physical and virtualized environmentsacross data centers, desktops, and other devices.In environments using this systems managementsuite, administrators can take advantage of theDell OpenManage Integration Suite for MicrosoftSystem Center to incorporate Dell hardwaremonitoring and management functionalityinto System Center.The Dell OpenManage Integration Suiteencompasses a comprehensive set of DellManagement Packs (MPs) and other toolsdesigned to simplify and automate essentialIT tasks, including the Dell Server MP Suite forDell PowerEdge and PowerVault servers, DellRemote Access Controllers (DRACs) and IntegratedDRACs (iDRACs), and Dell Chassis ManagementControllers (CMCs); the Dell Client MP for Dellbusiness client systems; the Dell Printer MP forDell printers; the Dell PowerVault MD StorageArrays MP Suite for Dell PowerVault MD Seriesstorage arrays; and the Dell EqualLogic Storage MPSuite for Dell EqualLogic PS Series Internet SCSI(iSCSI) storage area network (SAN) arrays. 1 Availableas software downloads at no additional cost,these tools integrate directly with System Centerapplications to deliver end-to-end managementand monitoring functionality for supportedDell hardware. Through the Dell OpenManage1 For more information, see “Unified Monitoring and Proactive Management of Dell Hardware with Microsoft System Center,” by ViswanathanBalakrishnan, Saravan Kumar, Mahendran P., and Vignesh Pandian, in Dell Power Solutions, 2010 Issue 1, dell.com/downloads/global/power/ps1q10-20100177-balakrishnan.pdf.56 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Integration Suite, administrators can select anddownload individual tools as needed, withouthaving to purchase a prescribed package oftools that may not be necessary in their specificenvironment.As part of the Dell OpenManage IntegrationSuite for Microsoft System Center, the DellEqualLogic Storage MP Suite enables ITadministrators to discover, classify, inventory, andmonitor Dell EqualLogic PS Series arrays—includingthe latest 10 Gigabit Ethernet EqualLogic PS6010and EqualLogic PS6510 arrays—using System CenterOperations Manager (SCOM) or System CenterEssentials (SCE). SCOM is typically suitable for usein large, enterprise-class environments, while SCE2010 is designed for midsize organizations with upto 50 servers and 500 client systems and providesa unified console for managing physical and virtualservers, clients, network devices, software, andIT services; in this article, Operations Manager isused to refer generically to both SCOM and SCE.In conjunction with System Center and the othertools in the Dell OpenManage Integration Suite,the EqualLogic Storage MP Suite offers a way foradministrators to incorporate their EqualLogicarrays into a comprehensive framework forefficient IT management. 2Dell EqualLogic Storage MP Suiteinstallation and architectureThe Dell EqualLogic MP installer is designed toperform all necessary installation tasks other thanimporting the MP into the Operations Managermanagement server, helping simplify initialdeployment for IT administrators. This processincludes installing the internal Dell EqualLogicHelper utility, which the MP uses to discoverand monitor EqualLogic PS Series arrays. TheEqualLogic Helper files are not installed onMicrosoft Windows ® client operating systems.After the installation is complete, administratorscan customize the MP using overrides availablefor various unit monitors for EqualLogic groupdiscovery and component-level monitoring to helpmeet their specific needs and data center policies.SCOM management serverDell EqualLogicMPDell EqualLogicHelper utilityDell EqualLogic storage groupFigure 1. Example Dell EqualLogic Storage MP Suite architecture in a large-scale data centerFigure 1 shows an example EqualLogic StorageMP Suite architecture in a large-scale data center.This infrastructure would comprise a SCOM rootmanagement server (RMS); device-specific SCOMmanagement servers that monitor various Delldevices, including a dedicated, device-specificmanagement server for monitoring the EqualLogicarrays; a SCOM database server; and a SCOM datawarehouse server.After the EqualLogic MP has been importedinto Operations Manager, administrators candiscover the EqualLogic arrays by performing anout-of-band Simple Network Management Protocol(SNMP) v2c–based discovery using the EqualLogicmanagement port IP address (which may be eitherthe group IP address or the dedicated managementport IP address). Operations Manager can thenmonitor the SNMP-based hardware traps of thediscovered EqualLogic group and classify the arraysin the Operations Manager alert view, enablingadministrators to proactively view and managethe resulting hardware alerts. The MP provides aseparate view in the Operations Manager consoleshowing alerts related to possible impendingfailures, including cause and resolution informationto help administrators troubleshoot the arraysby launching the EqualLogic console and/orphysically accessing the array. The MP also verifiesthat discovered arrays have firmware versions of4.x.x or higher, and generates alerts listing the IPaddresses of groups with members that do notmeet this requirement.As part of the MP monitoring functionality,monitored arrays are represented with variousSCOM managementgroup RMSImported Dell MPsSCOMdatabaseserverSCOM datawarehouseserver2 The Dell EqualLogic Storage MP Suite is available for download at support.us.dell.com/support/downloads/format.aspx?releaseid=R266477. For a list ofsupported EqualLogic arrays and detailed functionality provided by the EqualLogic MP, see the documentation available at support.dell.com/support/edocs/software/smconect/msscom/40eqlmp.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 57

Storage optimizationSimplifying storagemonitoring and managementDell EqualLogic PS Series arrays provide the advantages of consolidatednetworked storage in a self-managing Internet SCSI (iSCSI) storage areanetwork (SAN) designed for simplicity and cost-effectiveness even in large-scaledeployments. By helping to eliminate complex tasks and enabling fast, flexiblestorage provisioning, these arrays can help dramatically reduce the costs ofstorage acquisition and ongoing operations. These fault-tolerant, fully redundantarrays include software features enabling administrators to virtualize, optimize,and protect their data, while avoiding the management hassles that can comewith multiple software licenses or support contracts. Organizations can also adda yearly support contract that enables them to implement new software featuresand enhancements as they become available, extending the value of theirhardware investment.In environments that include EqualLogic arrays, administrators can takeadvantage of a comprehensive suite of software options to help simplify storagemanagement, enable advanced functions, and support virtualized storage andbest-practice data protection, including the following tools:EqualLogic PS Group Manager: This simple, intuitive, unified storagemanagement and provisioning tool allows administrators to remotely managevirtually any aspect of their EqualLogic SANs using a Web browser–basedinterface, helping eliminate the need for a dedicated management workstationor server. In addition to using this graphical interface, administrators can managearrays using a scriptable command-line interface over Secure Shell (SSH) andTelnet. Standard features include built-in monitoring and notifications throughe-mail, syslog support, and comprehensive Simple Network ManagementProtocol (SNMP) monitoring and traps.Enclosure Monitoring System (EMS): EMS provides the real-time status ofEqualLogic array components such as power supplies, controllers, fan trays, diskdrives, and network interface ports. It allows administrators to quickly recognize,report, and address component faults and failures without downtime, helpingmaintain system availability.SAN HeadQuarters (SAN HQ): Administrators can use SAN HQ to monitorEqualLogic arrays, pools, and groups from a single location. This centralizedtool supports performance, alarm, and health monitoring as well as historicalreporting to assist with operational planning, trend analysis, and troubleshooting.Flexible viewing options range from high-level group summaries to sophisticatedanalyses, including detailed statistics on performance, latency, capacity, volumes,and network activity. Predefined and customized views enable administrators toselect their preferred method of accessing information.components, including a Controllers group forcontrollers, a Physical Disks group for hard disks, aNetwork Interfaces group for Ethernet interfaces,a Power Supply group for power supplies, anda Sensors group for temperatures and fans. Thisfeature enables administrators to monitor anassortment of EqualLogic array components asindividual field replaceable units (FRUs). In addition,the diagram view shows a Volumes group thatincludes the volume information associated withthe arrays, enabling administrators to narrow downspecific impending hardware failures and takeappropriate action. Components in the diagramview are represented hierarchically for specificEqualLogic groups under the common DellHardware group, which includes the global healthstatus of all Dell hardware presented by the baseDell MP. By default, the MP’s component-leveldiscovery cycle is set to one day and the healthmonitoring cycle is set to approximately one hour;administrators can customize these periods usingconfiguration overrides.The EqualLogic PS Group Manager tool is anatural extension of the EqualLogic MP, in that itenables storage administrators to easily provisionstorage, schedule replication, and carry outcomprehensive array management, including criticalhardware components of the EqualLogic arrays. Akey feature within Group Manager is the EnclosureMonitoring System (EMS), which tracks the healthof redundant components such as power supplies,controllers, fan trays, disk drives, and networkinterface ports. Leveraging SNMP-based hardwarealerts, EMS through Group Manager can notifyadministrators that a component has reached a“warning” or “critical” level, thus enabling a proactiveapproach to managing component failures. (Formore information on EMS and other EqualLogicmanagement tools, see the “Simplifying storagemonitoring and management” sidebar in this article.)Flexible views for storage monitoringDifferent Dell MPs categorize Dell hardwareinto Dell-specific groups, which could includeDell EqualLogic PS Series storage arrays, DellPowerVault MD Series storage arrays, MicrosoftWindows–based Dell servers and clients, DRACs,CMCs, and Dell printers. Status monitoring for58 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Dell systems includes Dell-specific alerts andpre-failure alerts. Administrators can also launchspecific Dell instrumentation consoles from theOperations Manager console to perform granularproblem-solving analysis and enable effectivesystems management.The Dell EqualLogic MP presents the followingviews in Operations Manager:• Alert views: Alert views present specific alertsgenerated by Dell hardware. In the case of theEqualLogic MP, alerts are displayed for SNMPhardware events received from the monitoredarrays, along with cause and resolutioninformation to assist with troubleshooting (seeFigures 2 and 3). Administrators can also addcustom fields for group- and member-relatedinformation to help them track down alerts thathave come from a specific array.• Diagram views: Diagram views provide aholistic, graphical representation of hardwaredevices and logical device groups (see Figure 4).Administrators can further drill down thehierarchy to monitor arrays and other hardwareat a granular level. This view includes realistic,component-specific icons for various FRUcomponents in the EqualLogic arrays. In a datacenter environment where administrators areusing both the EqualLogic MP and the MD SeriesStorage Array MP, this view can also visuallyclassify EqualLogic and PowerVault MD Seriesstorage arrays using distinctive icons to helpadministrators easily identify these arrays.• State views: State views display health statusbased on the event and status poll informationreported by the hardware (see Figure 5). TheEqualLogic MP provides a list of severity-levelindicators along with detailed inventoryinformation to help administrators monitor andmanage their arrays. By default, the OperationsManager alert mapping is “OK,” “warning,” and“critical”; the various states of the hardware alertsfrom the EqualLogic array, like “unknown” and“warning,” are modeled as “warning” because ofthe Operations Manager state representation.• Detail views: Within the alert, diagram, and stateviews, administrators can select an individualEqualLogic group to view specific informationFigure 2. Alert view for Dell EqualLogic arrays in Microsoft System Center Operations Manager 2007 R2Dell EqualLogic alert name1 RAID set double faulted Warning2 One fan tray removed Warning3 RAID lost block table full Warning4 Battery has less than 72 hours of charge Warning5 RAID orphan cache Warning6 Multiple RAID sets found Warning7 Incorrect controller module inserted Warning8 Low ambient temperature Warning9 Ops panel missing or broken Warning10 Enclosure is open for a long time Warning11 EqualLogic disk status changes from one state to another Informational12 Temperature sensor high threshold exceeded Critical13 Temperature sensor low threshold exceeded Critical14 Fan speed high threshold exceeded Critical15 Fan speed low threshold exceeded Critical16 Power supply fan failure Critical17 Power supply failure Critical18 Both fan trays removed Critical19 RAID lost cache Critical20 NVRAM battery failed Critical21 Critical hardware component failed Critical22 EMM link failure Critical23 High battery temperature Critical24 Both the Sumo channel cards missing Critical25 EIP failed in Sumo Critical26 Both the Sumo channel cards failed CriticalFigure 3. SNMP-based Dell EqualLogic hardware alerts and severity definitionsDell EqualLogic MP severitydefinition in Operations ManagerReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 59

Storage optimizationon that group and its members. The componentattributes in these detail views have beenselected to provide information that wouldbe useful to administrators from a hardwaremanagement perspective.To provide comprehensive details onmonitored arrays and enable administrators totroubleshoot their arrays, the EqualLogic MP alsoincludes a Launch EqualLogic Console task withinOperations Manager that administrators can useto launch the Web browser–based EqualLogic PSGroup Manager console.Figure 4. Diagram view of Dell EqualLogic arrays and other hardware in Microsoft System CenterOperations Manager 2007 R2Unified monitoring andmanagement infrastructureMicrosoft System Center is designed to supporta comprehensive hardware monitoring andmanagement infrastructure. As part of the DellOpenManage Integration Suite for MicrosoftSystem Center, the Dell EqualLogic Storage MPSuite enables administrators to incorporate theirEqualLogic PS Series iSCSI SAN arrays into a SCOMor SCE management server console alongsideother Dell hardware. Taking advantage of thisunified approach can help IT departments respondquickly to hardware alerts while simplifyingmanagement of the environment as a whole.Viswanathan Balakrishnan is a softwarevalidation lead engineer on the Dell BusinessSoftware Validation team specializing in enterpriseand client systems management and virtualization.Stanley L. Stevens is a virtualization solutionsmarketing manager in the Dell Large EnterpriseStorage Marketing Group.Sekhar Duggirala is a software tester on the DellEnterprise Software Validation team with a specialfocus on Microsoft System Center.Figure 5. State view of Dell EqualLogic arrays in Microsoft System Center OperationsManager 2007 R2Tom George is a software development engineeranalyst on the Dell Partner Engineering Teaminvolved in integration development for MicrosoftSystem Center.Learn moreDell and MicrosoftSystem Center:dell.com/systemcenterDell OpenManage:dell.com/openmanageDell EqualLogic PS Series:dell.com/psseriesdell.com/equallogicMicrosoft System Center:microsoft.com/systemcenter60 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Storage optimizationBoost storage consolidationwith Blue Coat WANOptimization and DellPowerVault appliancesBy George Sadler and Suresh JasrasariaDeduplication, compression, and bandwidth enhancementtechnologies available in Dell PowerVault DL2100 backupand Blue Coat ® ProxySG ® WAN Optimization appliancescan alleviate stress on networks—helping organizationsenhance data protection and application recovery at remoteor branch offices.Demand for increased storage, fueledby exponential data growth, can resultin storage sprawl that can be difficultto manage at both remote or branchoffice (ROBO) and central data center locations.ROBO locations, in particular, often do not havefull-time IT personnel, and the staff may not haveexpertise in managing storage and performingbackup and recovery operations. These constraintscan put ROBO data particularly at risk of loss ornoncompliance with regulatory requirements.Storage consolidation can be a key strategyfor overcoming data protection and managementchallenges at ROBO locations. Consolidation enablesadministrators to store data at ROBOs on a singlesystem, and storage can then be replicated to acentralized repository for data protection, retention,and analysis—helping alleviate the need for ROBOIT personnel to perform complicated data protectionand management tasks.Although storage consolidation at ROBOlocations can offer many benefits, the cost andavailability of high-bandwidth lines coupled with highlatency in wide area network (WAN) connectivitycan present serious challenges to a storageconsolidation strategy for ROBOs. Deploying limitedbandwidth between a ROBO and a central datacenter can constrain storage consolidation at theROBO—for example, the amount of time it takesto replicate consolidated data from the ROBO tothe central location can be larger than the recoverypoint objective (RPO) of the organization, and theamount of time it takes to restore consolidatedROBO data from the central data center may belarger than the recovery time objective (RTO) ofthe organization, thus jeopardizing applicationrecoverability at the ROBO. Such WAN constraintscan be especially pronounced in ROBOs with alarge number of users or high data change rates.Fortunately, emerging technologies such as datadeduplication, data compression, and bandwidthoptimization can alleviate stress on WANs,enabling organizations to boost ROBO storageconsolidation with reduced WAN traffic.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 61

Storage optimizationDell EqualLogic PS4000XiSCSI SAN arrayBlue Coat ProxySG Virtual Appliance(installed on Dell PowerEdge 2950 server)Dell PowerConnect 6248switchCisco Catalyst 3560 switchDell PowerEdge R710serverWANBlue Coat ProxySG 510applianceDell PowerConnect 5448switchFSCT clientsFSCT controllerDell PowerConnect 6248switchDell EqualLogic PS6000EiSCSI SAN arrayFigure 1. Test configuration for the low-change-rate environmentTo help organizations implement a successful,scalable ROBO storage consolidation strategy, Delland its partners offer a range of storage arrays andappliances—including Dell EqualLogic PS SeriesInternet SCSI (iSCSI) storage area network (SAN)arrays, Dell PowerVault DL2100 backup appliances,and Blue Coat ProxySG WAN Optimizationappliances. EqualLogic arrays enable storageconsolidation at remote sites and support automatedreplication to a centralized location, providing anexcellent foundation for implementing a storageconsolidation strategy. PowerVault DL2100 backupappliances offer built-in data deduplication andcompression features that help to significantly reduceWAN traffic and provide reliable, fast local recovery atremote sites. And Blue Coat ProxySG appliances offeradvanced features for optimizing WAN connectivity,including data deduplication, data compression,and network protocol enhancements designed tofurther reduce WAN traffic to meet an organization’sRPO and RTO objectives with minimum networkbandwidth. In fact, benchmark testing has shownthat adding a Blue Coat ProxySG WAN Optimizationappliance in an existing configuration can enableROBOs to support up to five times more users(see the “Benchmarking replication and backupperformance” section in this article).Organizations can take advantage of theseappliances to both enhance data protectionand application recovery through storageconsolidation, and overcome WAN constraintsbecause of limited-bandwidth deploymentsbetween a ROBO and a central office. Byimplementing a storage consolidation strategybased on EqualLogic PS Series storage arrays,PowerVault DL2100 backup appliances, and BlueCoat ProxySG WAN Optimization appliances,they can create reliable, scalable storageconsolidation and data protection and recoveryfor their ROBOs—enabling them to streamline datamanagement, ensure regulatory compliance, andmeet organizational RPO and RTO requirements.Designing a consolidatedstorage infrastructureA storage consolidation strategy based on DellEqualLogic PS Series iSCSI SAN arrays helps removethe burden of data protection and regulatorycompliance from ROBOs. Implementing an appropriateconfiguration, however, depends on factors such asnumber of remote users or data change rates, andavailable WAN bandwidth and latency.ROBOs with a low data change rate or a smallnumber of users as well as a connection to thecentral data center with adequate bandwidthand latency may be able to use a storageconsolidation strategy based on EqualLogicarrays alone, using the built-in remote replicationfunctionality included at no additional cost. Atypical deployment might include EqualLogicPS4000 series arrays for storage consolidationat ROBO locations connected to a central data62 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

center approximately 2,000 miles away (40 msround-trip latency) over a 5 Mbps connection. Insuch a deployment, the ROBO EqualLogic PS4000array would replicate the changed data every hour,a typical RPO objective, to an EqualLogic PS6000series array at the central data center for dataprotection, retention, and analysis.ROBOs with a high data change rate or a largenumber of users with a limited-bandwidth connectionto a central data center—or organizations thatrequire extremely fast, local disaster recovery (verysmall RTO)—have the option of using a storageconsolidation strategy that combines EqualLogic PSSeries arrays with Dell PowerVault DL2100 backupappliances. PowerVault DL2100 backup appliancesinclude built-in compression and data deduplicationfeatures designed to avoid backup of redundantfiles, attachments, and documents and help reducebackup storage requirements—which in turn candramatically reduce WAN traffic associated withremote data protection and recovery. A typicaldeployment for such ROBO locations mightinclude EqualLogic PS4000 series arrays for storageconsolidation with PowerVault DL2100 backupappliances for local backup and quick recovery. Insuch a deployment, the ROBO PowerVault DL2100backup appliance would deduplicate, compress, andreplicate the changed data every hour, a typical RPOobjective, to an additional PowerVault DL2100 backupappliance at the central site over a limited-bandwidthconnection (for example, a 5 Mbps connection witha 40 ms round-trip latency) for data protection usingCommVault ® Simpana ® backup software.Optimizing WAN performanceTo further enhance the effectiveness of WANconnections associated with data protection andrecovery, or to help scale ROBO data growthand consolidation, organizations can add BlueCoat WAN Optimization. Blue Coat ProxySGappliances—available as either physical (hardware)or virtual (software running in a dedicated server orvirtual machine) implementations—use a specificcombination of the following technologies toenhance WAN effectiveness:symbolizing those patterns as tokens and thensending tokens in lieu of bulky traffic• Object caching: Can lighten network trafficeven further by caching entire applicationspecificobjects• In-line compression: Complements caching byhelping to reduce data over the wire• Protocol optimization: Can utilize a variety ofoptimization techniques to enhance the efficiencyof WAN network protocols such as TCP, iSCSI,HTTP, HTTP over Secure Sockets Layer (HTTPS),and Common Internet File System (CIFS)• Bandwidth management: Helps partitionapplication traffic and allocate bandwidth toapplications based on priorityAdding a Blue Coat ProxySG appliance to astorage consolidation configuration based on DellEqualLogic arrays alone or on Dell EqualLogicarrays combined with Dell PowerVault DL2100backup appliances can help dramatically increasebackup and recovery performance over existingWAN connections. Enhanced WAN effectiveness, inturn, enables organizations to scale ROBOs withoutnecessarily having to purchase additional WANbandwidth. In addition, when ROBOs do not uselocal storage and run critical business applicationslike e-mail or enterprise resource planning (ERP)software over a WAN using remote desktops, aBlue Coat ProxySG appliance can help increaseapplication performance while reducing WAN costs.Benchmarking replicationand backup performanceTo illustrate the benefits of adding Blue CoatWAN Optimization to a storage consolidationHourly data transfer (GB)7654321Blue CoataccelerationdemoThis video overviewof Blue Coat WANOptimization includesa detailed discussion ofspecific use cases as wellas demonstrations ofthe Blue Coat dashboardinterface.dc.bluecoat.com/?elqPURLPage=510Figure 2. Replication results for thelow-change-rate environment withand without Blue Coat ProxySG WANOptimization• Byte caching: Helps reduce network traffic byobserving repetitive patterns in network traffic,0Without Blue CoatProxySG VirtualApplianceWith Blue CoatProxySG VirtualApplianceReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 63

Storage optimizationdeployment, in April 2010 PrincipledTechnologies conducted Dell-commissionedbenchmark tests on various ROBOconfigurations. 1 In particular, these testswere designed to determine the advantagesof various remote replication and backuptechnologies for ROBO data protection to acentral office across a WAN. Configurationswith and without a Blue Coat ProxySGphysical appliance and a Blue Coat ProxySGVirtual Appliance were tested.These tests used the Microsoft ® FileServer Capacity Tool (FSCT) with theHomeFolders workload to simulate typicalROBO data change rates in two exampleconfigurations—one based on a low datachange rate and the other based on a highdata change rate. FSCT simulates file serveractivity, and includes typical operationssuch as deleting, navigating, and movingfiles using Windows ® Explorer and editing,saving, closing, and opening files withMicrosoft Office software such as the Word,Excel ® , and PowerPoint ® applications. Atypical WAN bandwidth of 5 Mbps anda ROBO at a distance of approximately2,000 miles (round-trip delay of 40 ms)were used in all tests.Dell EqualLogicPS4000XiSCSI SAN arrayDell PowerConnect6248 switchDell PowerEdgeR710 serverDell PowerConnect5448 switchFSCT clientsFSCT controllerWANFigure 3. Test configuration for the high-change-rate environmentLow data change rateThe test configuration illustrated in Figure 1simulated a ROBO with a low data changerate. This ROBO configuration consisted ofone Dell PowerEdge R710 rack server forfile sharing, one Dell EqualLogic PS4000XiSCSI SAN array for primary storage and localbackup, and one Blue Coat ProxySG VirtualAppliance installed on a PowerEdge 2950server for WAN optimization. A CiscoLayer 3 switch was used for redirecting trafficto the Blue Coat ProxySG Virtual Applianceusing Web Cache Communication Protocol(WCCP). The central office configurationconsisted of one Blue Coat ProxySG 510hardware appliance for WAN optimizationand one Dell EqualLogic PS6000E array foroff-site data protection.The test first ran the FSCT benchmarkat the ROBO with 150 users for fourhours to create a baseline data set onthe EqualLogic PS4000X array. Thisbaseline data set was then replicatedto the EqualLogic PS6000E array usingEqualLogic replication. To evaluate ROBOdata protection performance for a range ofusers, the test then restarted and ran FSCTwith user counts ranging from 10 to 150,Dell PowerVaultDL2100 applianceBlue CoatProxySG 510applianceBlue CoatProxySG 510applianceDell PowerConnect6248 switchDell PowerVaultDL2100 applianceincreasing the user count progressivelyevery other hour.Backups were performed every hour, atypical RPO for ROBOs, where the amountof data replicated to the central locationeach hour was equal to the amount ofdata that changed in the hour precedingthe backup and replication. The tests wereperformed both with and without the BlueCoat ProxySG Virtual Appliance.Figure 2 shows the results. Theconfiguration without the Blue Coat ProxySGVirtual Appliance was able to support up to1.85 GB of user changes in less than onehour, while the configuration with the BlueCoat ProxySG Virtual Appliance was able tosupport up to 6.07 GB of user changes inless than one hour—a more than threefoldincrease in effective network bandwidth dueto Blue Coat WAN Optimization.High data change rateThis test was designed to illustrate WANoptimization in a ROBO with a high datachange rate when using a Blue Coat ProxySG510 hardware appliance (see Figure 3). TheROBO configuration consisted of one DellPowerEdge R710 rack server for file sharing,one Dell EqualLogic PS4000X array for primarystorage, one Dell PowerVault DL2100 backupappliance for local backup, and one Blue CoatProxySG 510 hardware appliance for WANoptimization. The central office configurationconsisted of one Dell PowerVault DL2100backup appliance for off-site data protectionand one Blue Coat ProxySG 510 hardwareappliance for WAN optimization.The test first ran the FSCT benchmark atthe ROBO with 500 users for four hours tocreate a baseline data set on the EqualLogicPS4000X array. This baseline data setwas backed up to the remote PowerVaultDL2100 appliance and then replicated to thecentralized PowerVault DL2100 applianceusing CommVault Simpana backup softwarewith continuous data replication. To evaluate1 For the complete report, including detailed information on the test environment, benchmark workloads, methodology, and results, see “Dell PowerVault DL2100 and Dell EqualLogic PS Series RemoteOffice/Branch Office Data Replication with Blue Coat ProxySG WAN Optimization,” by Principled Technologies, April 2010, principledtechnologies.com/clients/reports/dell/dl2100robowithbluecoat.pdf.64 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

ROBO data protection performance for a rangeof users, the test then restarted and ran FSCT withuser counts ranging from 30 to 500, increasing theuser count progressively every other hour.As in the low-change-rate test, backups wereperformed every hour, and the amount of datareplicated to the central location each hour wasequal to the amount of data that changed in thehour preceding the backup and replication. The testswere performed both with and without the Blue CoatProxySG 510 appliances.Figure 4 shows the results. The configurationwithout the Blue Coat ProxySG 510 appliances wasable to support up to 1.75 GB of user changes inless than one hour, while the configuration withthe Blue Coat ProxySG 510 appliances was ableto support up to 9.19 GB of user changes in lessthan one hour—a more than fivefold increase ineffective network bandwidth due to Blue CoatWAN Optimization.Boosting storage consolidation andscalability for remote or branch officesA storage consolidation strategy based on DellEqualLogic PS Series iSCSI SAN arrays helpsorganizations to streamline data managementand ensure reliable data protection andrecovery. The powerful deduplication,compression, and network optimizationtechnologies available with Dell PowerVaultDL2100 backup appliances and Blue CoatProxySG WAN Optimization appliances helpHourly data transfer (GB)109876543210organizations overcome WAN bandwidthlimitations to enhance data protection andapplication recovery. By boosting storageconsolidation scalability in this manner,organizations can enhance the performance oftheir data protection and recovery at ROBOs tohelp facilitate regulatory compliance, streamlineIT management, and meet RPO and RTOservice-level agreements.Learn moreWithout Blue CoatProxySG 510applianceWith Blue CoatProxySG 510applianceDell EqualLogic PS Series:dell.com/psseriesdell.com/equallogicBlue Coat ProxySG:bluecoat.com/products/sgFigure 4. Replication resultsfor the high-change-rateenvironment with and withoutBlue Coat ProxySG WANOptimizationGeorge Sadler is a verticalsolutions marketingstrategist for the DellEnterprise SolutionsMarketing team, andhas more than 20 yearsof experience spanningmultiple industries.Suresh Jasrasaria is aproduct marketing seniorconsultant in the DellEnterprise Storage ProductGroup, and has worked inthe data storage industry formore than 20 years.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.Backup 2.0 solutions from Vizioncore provide simply better data protection for all environments.They use images to deliver better backup, transmission, and recovery for physicaland virtual systems.Simply BetterData ProtectionBackup 2.0:Gain $15 Net Savingsfor Every $1 InvestedVizioncore provides:· Image-based backup and recovery of virtual machines (VM) that eliminates agent costs and optimizes backupwindows· “Backup once, restore anywhere” allows admins to use a single backup image to recover entire system state,individual files and folders, and application-level objects· Image-based replication that reduces recovery time objectives (RTO) and recovery point objectives (RPO) thatis easy to use and affordable· Complete system protection with full system recovery, file-level restore, and object-level recoverywww.vizioncore.com/backup20

Storage optimizationF5 ARX file virtualizationand Dell storageBy Renny ShenF5 ® ARX ® Series intelligent file virtualization devices andDell storage help simplify management by enablingflexible data mobility, while automated data managementpolicies help to reduce costs, lower IT overhead, andenhance productivity.File virtualizationat the DellTechCenterThis post from the DellTechCenter wiki gives aquick rundown on theadvantages of F5 ARX filevirtualization in enterpriseIT environments—includingreduced storage costs,optimized backups,and nondisruptive datamigrations.delltechcenter.com/page/file+virtualization+with+f5The torrid rate of growth in file data iselevating the strategic importance ofsmart data management in the eyesof IT professionals. Businesses andother organizations are generating data fasterand retaining it for longer than ever—but in manyenvironments, the traditional strategy of constantlyadding capacity no longer scales well enoughto handle this accelerated growth. Bridging thegap between rapid data growth and constrainedIT budgets, Dell and F5 Networks have partneredto help organizations build dynamic storageinfrastructures that can simplify management andoffer breakthrough storage economics.File virtualization using F5 ARX Series devicesprovides the foundation for flexible data mobility,while automated management policies intelligentlymove files based on data value. This combination ofnondisruptive file movement and automation canhelp organizations make the most of their investmentin—and leverage the cost and managementadvantages of—Dell file storage systems.Understanding the impactof data growth on storageData growth stresses the storage infrastructurein many ways, but the capacity cost required tosustain that growth is generally paid the greatestattention. Storage is often the largest line item inthe IT budget, and the rate of data growth canexceed that of budget growth, often dramatically.But behind the balance sheet, the lack of datamobility can contribute as much to rising storagecosts as data growth. Moving data is a disruptivetask, making it difficult to match data to the mostcost-effective or efficient location.One common attribute of many high-growthenvironments is the high ratio of inactive datato active data. As organizations retain data forextended periods, a growing percentage of it is nolonger actively accessed or modified. The logicalnext step would be to move that inactive datato lower-cost storage, but the operational costof doing so can be prohibitive. Aside from the ITresources needed to continually identify whichfiles are inactive, the act of physically moving thosefiles can be disruptive: when users and applicationsare statically mapped to specific storage resources,moving data to a different resource can breakthose mappings and cause downtime.Another critical, yet often overlooked, aspectof data growth is its inconsistent nature. Becausedata sets grow at different rates and those ratestypically change over time, it is extremely difficult66 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

“F5 ARX Series devices provide a rich setof automated policies that govern themovement of individual files customized tohow specific organizations intend to deployand optimize their available storage capacity.”to predict which storage resources will needadditional capacity and which will have excesscapacity. Here, once again, a lack of data mobilityhampers effective growth management by makingit hard to share capacity across storage resourcesor to migrate data between resources to balanceutilization levels. As a result, organizations are oftenforced to deploy more capacity than they reallyneed, while existing capacity goes unclaimed.Enabling data mobilityF5 ARX Series intelligent file virtualization devicescan play a crucial role in addressing thesedata mobility challenges. File virtualization isdesigned to abstract the physical storage layerand decouple the physical relationship betweenclients and storage resources. The key featurethat facilitates this decoupling is the globalnamespace—essentially a collection of virtualCommon Internet File System (CIFS) shares andNetwork File System (NFS) exports presented bythe ARX device. With file virtualization, clientsmount these logical or virtual shares and exportsto access their data, instead of mapping to thephysical storage devices directly.Because user and application clients mountvirtual shares and exports, they are no longerexposed to the physical location of their data.This approach enables ARX devices to move databetween different physical locations withoutchanging how clients logically access data. In asense, the device acts like a file router, keepingtrack of the current location of each file androuting logical access through virtual shares to theappropriate physical location.The in-line placement of ARX devicesbetween client systems and storage also makesthem well suited for providing and enforcingdata management policies. The devices providea rich set of automated policies that governthe movement of individual files customized tohow specific organizations intend to deploy andoptimize their available storage capacity, includingthe following:• Data migrations: Migrations occur for avariety of reasons, and can include plannedevents such as consolidations, upgrades, andtechnology refreshes as well as unplannedevents such as reprovisioning capacity.Virtualized file storage enables organizations toflexibly move data as needed without affectinguser or application access. Data migrationpolicies help reduce the amount of up-frontTier 1Optimized for performance(SAS or SSD disks)Active dataUser and application clientsF5 ARX SeriesIntelligent file virtualizationTier 2Optimized for capacity(SATA disks)Online file migrationbetween heterogeneous tiersTier 3Optimized forlong-term archiving(deduplicated SATA disks)Inactive dataFigure 1. Automated storage tiering with F5 ARX Series intelligent file virtualization and Dell storageReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 67

Storage optimization“F5 ARX Series devices and F5Data Manager software can helporganizations realize the cost andmanagement advantages of Dellfile storage systems.”planning, time, and IT overhead required, evenwhen performing complex migrations andrestructuring file system layouts.• Storage tiering: Storage tiering helpsorganizations align the cost of storing data withtheir limited IT budgets. The basic premise is tomatch different classes of data with the mostappropriate type of storage. Many organizationshave already implemented tiering at the blocklevel using storage such as Dell EqualLogic PS Series Internet SCSI (iSCSI) storage areanetwork (SAN) arrays, or on a per-applicationbasis. ARX devices provide automated storagetiering policies that take advantage of theinherent business context of each file to helpdetermine the most appropriate place tostore it—active or high-value data on highperformancestorage, and inactive or low-valuefiles on lower-cost storage. When inactivedata represents a large percentage of all dataunder management, storage tiering can bea huge source of cost savings. In addition,organizations can now apply differentbackup policies to tiers with active andinactive data to help reduce backup windowsand media consumption.• Capacity balancing: Capacity balancingcombines multiple file storage resources intoa virtual storage pool. Similar to the capabilitiesof Dell EqualLogic PS Series storage betweendifferent arrays at the block level, virtual filesystems presented through ARX devices canutilize storage capacity from different back-endfile storage resources—allowing organizationsto reclaim stranded capacity from underutilizedstorage and redeploy it where data is growing.Automated policies can balance file placementacross multiple storage resources for consistentaggregate utilization.Although the lack of data mobility hampersthe ability of organizations to respond to datagrowth, enabling data mobility by itself is notenough to do so effectively. Effective datamanagement must be a repetitive process ofoptimizing how and where storage capacity ismost efficiently and cost-effectively deployed.F5 Data Manager software can helporganizations understand their changing datastorage requirements by providing customizedreporting on file storage and data characteristics.Detailed analysis of scanned data providescrucial information for planning and refining datamanagement policies.Taking advantage of Dell file storageF5 ARX Series devices and F5 Data Managersoftware can help organizations realize the costand management advantages of Dell file storagesystems. Organizations looking to migrate theirstorage infrastructure to Dell systems can doso without disruption or downtime, while thoselooking to augment their existing environmentcan take advantage of ARX automated tiering andload balancing policies to seamlessly integratethese systems:• Seamless integration of Dell storage: ARXdevices are designed to simplify managementof heterogeneous file storage environmentsand allow organizations to deploy Dell filestorage sys tems seamlessly alongside existingsystems. File virtualization masks the physicaldifferences between different storage typesfrom user and application clients, enablingorganizations to utilize available storagecapacity efficiently and effectively regardlessof type, platform, or vendor.• Simplified migration to Dell storage: ARXdevices help simplify and accelerate datamigration projects from legacy infrastructuresto Dell file storage systems. One of the biggesthurdles when replacing legacy file storage isthe potential disruption from migrating data.68 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

ARX devices support online migration withoutdowntime or client reconfiguration, helpingto reduce planning, migration time, and IToverhead requirements.• Consolidation with Dell NX4 systems:ARX devices can perform nondisruptiveconsolidation of legacy file servers ontoDell NX4 systems for mixed Microsoft ®Windows ® , Linux ® , and UNIX ® environments.File virtualization allows nondisruptivefile migration, enabling even complexconsolidations with reduced planning,migration time, and IT overhead requirements.• Storage tiering with Dell EqualLogic and DellPowerVault systems: Storage tiering withARX devices can provide performance andcost-savings benefits for file data similar tothose provided by EqualLogic arrays for blockdata. Within a virtual file system, ARX devicescan migrate individual files between physicalCIFS shares or NFS exports presented bymultiple PowerVault systems and provisionedfrom Serial ATA (SATA), Serial Attached SCSI(SAS), and solid-state drive (SSD) capacityon EqualLogic arrays. File-level tieringcomplements block-level tiering by addingdimensions or criteria used to determinewhat data to move and when to move it.Organizations can configure automatedpolicies based on the inherent context of eachfile, using a combination of file attributes suchas the latest modification or access date, type,name, location, and size.• Storage tiering with Dell/EMC DD Seriessystems: ARX devices can also be used to takeadvantage of the data deduplication capabilitiesof Dell/EMC DD Series systems in a long-termfile archive tier. Because of the performanceimpact that can be associated with accessingdeduplicated data, organizations may chooseto deploy deduplicated storage capacity onlyfor data that is rarely accessed or modified.Automated tiering policies can move individualfiles as they age to a deduplicated storagetier, allowing organizations to seamlesslycombine primary storage capacity fromDell NX4 or EqualLogic systems for activedata with deduplicated storage capacity fromDell/EMC DD Series systems for inactive data(see Figure 1).• Load balancing across Dell storage: ARX devicescan combine Dell file storage systems—includingDell NX4, Dell EqualLogic, Dell PowerVault, andDell/EMC DD Series systems—into a single virtualstorage pool. Organizations with applicationsneeding large workspaces can create virtualfile systems that exceed the physical limitationsof individual devices while automated policiesdistribute files across multiple heterogeneousstorage resources for consistent andpredictable utilization.• Simplified provisioning of Dell storage: ARXdevices help make it easy to add capacity fromDell file storage devices into existing virtualshares, on demand and transparently to clients.By avoiding the disruption of reprovisioningstorage, ARX devices provide organizations withthe operational flexibility to run highly efficientstorage environments that are responsive tounpredictable growth.Building a dynamic storage infrastructureF5 ARX Series intelligent file virtualizationdevices and Dell file storage systemscan help organizations build dynamicstorage infrastructures that can respond tounpredictable and inconsistent data growth.With F5 ARX Series devices, the combinationof data mobility and automated policies helpsdramatically simplify the processes of deployingand managing a comprehensive range of Dellfile storage systems while helping maximizeoperational and cost savings.Learn moreRenny Shen is a product marketing managerwith F5 Networks.Dell storage:dell.com/storageF5 ARX Series:f5.com/products/arx-seriesF5 Data Manager:f5.com/products/data-managerReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 69

Storage optimizationAccelerate and securedata replication withF5 WAN optimizationBy Fred Johnson, Ujjwal Rajbhandari, and Puneet DhawanCombining Dell EqualLogic PS Series storage with F5 ® BIG-IP ®application delivery systems and WAN Optimization Module technology can have significant performance and securitybenefits for data replication over wide area networks (WANs),helping reduce WAN-related costs.Application availability, data centercapacity, and business continuity canall depend on the timely movement ofapplications and data between remotesites. The low throughput and long data transfercompletion times typical of wide area networks(WANs) can make remote WAN replication difficultor even impossible. Additionally, the need to meetregulatory and compliance standards—includingcontrolled access to and encryption of replicateddata—is often a paramount consideration forenterprise storage administrators.With these challenges in mind, Dell and F5have completed proof-of-concept (POC) labtesting to demonstrate how organizations canhelp accelerate and secure Dell EqualLogic PSSeries Internet SCSI (iSCSI) storage area network(SAN) replication over WANs using version 10.1 ofthe F5 BIG-IP WAN Optimization Module (WOM).By overcoming the effects of limited bandwidth,high latency, and packet loss, WOM technologyhelps to dramatically increase throughput, reducereplication completion times, and enhancebandwidth utilization and efficiency. When takinginto account the costs of WAN bandwidth,equipment, and administration, these performanceadvantages can quickly lead to bottom-line savings.F5 BIG-IP WAN optimization featuresF5 BIG-IP application delivery controllers providethe flexibility to support multiple feature setson a single platform. For example, a BIG-IPdevice providing high availability, traditionalload balancing, and Secure Sockets Layer (SSL)offload can also incorporate WAN optimization,Web acceleration, application security, and otherfeatures without the need for additional dedicatedappliances—helping to simplify managementand save on hardware, rack space, and powerconsumption.F5 BIG-IP WOM technology is designed toaccelerate TCP traffic for data center applications,with the ability to scale to meet high bandwidthrequirements. The module includes specialacceleration profiles for Common InternetFile System (CIFS) and Messaging ApplicationProgramming Interface (MAPI), and supports WebCache Communication Protocol version 2 (WCCPv2);because it must be deployed symmetrically, itrequires two or more BIG-IP devices.70 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Figure 1. F5 BIG-IP WOM performance dashboardA variety of features helpto boost performance, securecommunications, and managetraffic, including the following:• iSessions: The F5 iSessionsnetwork tunneling featureintegrates security,acceleration, and trafficmanagement. This featureis designed to work throughfirewalls: for example, iSCSIrequires only TCP port 443 forthe secure control channeland TCP port 3260 for iSCSI,and iSessions encryption canbe enabled without changesto the port assignments.• Symmetric encryption:iSessions can be secured bycontrolling access, encryptingthe control channel, andoptionally encrypting theiSCSI replication traffic; theSSL encryption is designedto operate with only minimalimpact on performance.Administrators can easilyenable encryption bymodifying the iSCSI virtualserver using the Webbrowser–based BIG-IPmanagement interface.• Symmetric datadeduplication: In the WOMcontext, deduplication meanspreventing redundant datapatterns from crossing theWAN. A cache is built onthe device at each end, andwhen a duplicate pattern inthe network traffic is found, asmall reference to the cacheis transmitted instead of theentire pattern—an importantway to help reduce theamount of WAN traffic. TheDell and F5 POC testing useda memory deduplication datastore, an approach well suitedfor replication traffic. Thededuplication cache can alsobe stored on disk to supportincreased cache sizes.“The goal of WAN optimization is to increaseperformance by minimizing latency—helping toincrease the perceived bandwidth, seamlesslyrecover from packet loss, and enable WANs toprovide performance characteristics similar to LANs.”Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 71

Storage optimization• Symmetric adaptivecompression: As withsymmetric deduplication,compression helps to reducethe amount of WAN trafficand increase bandwidth usageefficiency. Adaptive means thatBIG-IP devices can intelligentlyselect compression based onthe type of data. Compressioncan also be offloaded tohardware compression cardson high-end F5 platforms tohelp reduce processor usageand increase throughput. TheDell and F5 POC testing usedadaptive compression, yieldinga combination of Lempel-Ziv-Oberhumer (LZO) and Deflate(level 1) compression. Volumedata set characteristics such ascompressibility influence theperformance gains associatedwith WOM technology.• TCP Express: The BIG-IPnetwork stack implements avariety of TCP optimizationtechniques and performancerelatedIETF Requestfor Comments (RFC)enhancements designed tooffload and optimize TCPnetwork traffic, helpingto reduce TCP overheadand increase networkperformance and reliability.• Layer 7 (L7) quality-ofservice(QoS) rate shaping:L7 QoS rate shaping allows forthe enforcement of bandwidthminimums and maximums perapplication, including burstcontrol, and supports termsof-serviceand differentiatedservices code point (DSCP)features. For example, thebandwidth associated with aniSCSI BIG-IP virtual server canbe limited, helping preventiSCSI replication from affectingother critical applications thatshare the same WAN link.The WOM performancedashboard offers integratedmonitoring and reporting tools,providing administrators witha quick look at real-time data,performance, and bandwidthgains for WOM-optimized traffic(see Figure 1).Proof-of-concepttest environmentThe goal of the Dell and F5 POCtesting was to demonstrate DellF5 BIG-IPconfigurationIn this Dell TechCenter video,Fred Johnson demonstrateshow to configure the F5BIG-IP WAN OptimizationModule for a central office.delltechcenter.com/video/9320147/f5+big+ip+configurationSimulated data center 1Dell PowerConnect6248 switchWAN emulatoriSessionsSimulated data center 2Dell PowerConnect6248 switchDell EqualLogicPS6000XViSCSI SAN arrayF5 BIG-IP 3600devices with WOMtechnologyF5 BIG-IP 3600devices with WOMtechnologyDell EqualLogicPS6000XViSCSI SAN arrayEncryption, compression, symmetric deduplication, and TCP optimizationsFigure 2. Proof-of-concept test environment incorporating F5 BIG-IP WOM technology72 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

800Throughput (MB/min)7006005004003002001m20s2m35s13m55s19m36s17m20s15m10s16m40sFigure 3. Dell EqualLogic replicationthroughput and completion timeswith F5 BIG-IP WOM accelerationover a 45 Mbps WAN with 100 ms oflatency and 1 percent packet loss1000500 MB 1 GB 5 GB 10 GB 10 GB 10 GB 10 GB(1:1) (1:1) (1:1) (1:1) (2:1) (4:1) (4:1 withencryption)Volume size and data compressibilityEqualLogic iSCSI SAN replicationthroughput and completiontimes when using F5 BIG-IPWOM technology to accelerateand encrypt iSCSI traffic as itcrossed a WAN. The tests wereperformed in November 2009at the Dell Interoperability Lab inRound Rock, Texas.As shown in Figure 2, thelab configuration included twoEqualLogic PS6000XV iSCSISAN arrays with sixteen 420 GB,15,000 rpm Serial Attached SCSI(SAS) hard drives in a RAID-50configuration connected bytwo Dell PowerConnect6248 L3 switches. F5 BIG-IP3600 appliances with WOMtechnology were connectedto the switches at eachsimulated site. A WAN emulatorinterconnected the switches;the emulator functions includedlimiting bandwidth, injectinglatency, and dropping packetsto simulate two remote datacenters connected over a WAN.The two arrays then replicatedvolumes over the emulatedWAN, benefiting from WOMacceleration and encryption.The test team used the Webbrowser–based EqualLogicGroup Manager user interfaceto configure the volumes,manage the replication jobs,and report throughput andcompletion times. The testruns included asynchronousfull-volume replication jobs, withno incremental replications, andeach run was limited to singlevolumereplications.Similarly, the Webbrowser–based BIG-IPmanagement interface wasused for WOM configuration,including endpoints, virtualservers, encryption profiles, andacceleration profiles and routing,as well as an iSCSI accelerationpolicy. The performancedashboard was monitored toobserve WAN optimizationstatistics, network throughput,and processor and memoryutilization during the test runs.F5 BIG-IP WOMacceleration test resultsThe goal of WAN optimizationis to increase performance byminimizing latency—helping toincrease the perceived bandwidth,seamlessly recover from packetloss, and enable WANs to provideperformance characteristicssimilar to LANs. Figure 3 showsDell EqualLogic iSCSI replicationthroughput and completion timeswith F5 BIG-IP WOM accelerationover a 45 Mbps WAN with 100 msof latency and 1 percent packetloss (10,000 packets per million)across a variety of volume sizes.The WOM-accelerated tests for the10 GB volume additionally showthe throughput and completiontimes for 2:1 and 4:1 volume datacompressibility as well as 4:1compressibility with encryption.To put these results incontext, using WOM technologyenabled the WAN link at 45 Mbpswith 100 ms of latency and1 percent packet loss to provideresults roughly comparable towhat could be expected of a100 Mbps LAN with no latencyor packet loss, even whenusing encryption—helping makeremote replication of large datavolumes possible in a muchshorter period of time than itwould otherwise require. As anReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 73

Storage optimization“F5 BIG-IP WOM technology can helporganizations reduce administrationcosts and avoid the need to purchaseadditional bandwidth while continuingto meet overall organizational andcompliance requirements.”Typical latency rangeTypical average latencyRegional within the United States or Europe 10–40 ms 30 msU.S. East Coast to West Coast 80–120 ms 100 msSouth America to North America 90–170 ms 150 msAsia Pacific to Europe 250–400 ms 300 msFigure 4. Typical latencies that might be expected across a variety of WAN link distancesadditional point of reference when evaluating theresults, Figure 4 shows typical latencies that mightbe expected across a variety of WAN link distances.Organizations can further increase iSCSI WANdata transfer throughput by configuring concurrentreplication of more than one volume at a time,helping maximize overall bandwidth utilization.They can also enhance network efficiency byusing remote point-in-time replication, which helpsminimize the amount of data crossing the WAN bytransferring only the changed volume data to thereplication partner.Reduced costs through WAN accelerationBecause bandwidth, equipment, andadministration for WAN operations can representsignificant costs over time, the performanceadvantages of F5 BIG-IP WOM technology withDell EqualLogic storage demonstrated by theDell and F5 POC tests can lead to bottom-linesavings. By accelerating data transfer timesover WANs, enabling encryption with minimalperformance overhead, and providing an easyto-usemanagement interface, F5 BIG-IP WOMtechnology can help organizations reduceadministration costs and avoid the need topurchase additional bandwidth while continuingto meet overall organizational and compliancerequirements. Combining WOM technologywith Dell EqualLogic iSCSI SAN arrays enablesorganizations to create a shared storage solutionsupporting accelerated and encrypted site-to-sitedata transfers over WANs.Learn moreFred Johnson is a strategic partner engineerat F5 Networks dedicated to Dell Labs.Ujjwal Rajbhandari is a product marketingconsultant for Dell storage.Puneet Dhawan is a technical marketingconsultant for Dell storage.F5 and Dell:f5.com/dellDell EqualLogic PS Series:dell.com/psseriesdell.com/equallogic74 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Storage optimizationA unified networking approachto iSCSI storage withBroadcom controllersBy Dhiraj Sehgal, Abhijit Aswath, and Srinivas ThodatiIn environments based on Internet SCSI (iSCSI) and10 Gigabit Ethernet, deploying Dell PowerEdge serverswith Broadcom ® NetXtreme II ® converged networkcontrollers can help IT departments to eliminate networkbottlenecks, reduce power consumption, and maximizedata center efficiency.Use of the Internet SCSI (iSCSI) protocolhas been growing significantly inenterprise data centers, largelybecause it offers a variety ofadvantages over traditional Fibre Channel storagetechnologies. Because iSCSI provides a simplemethod for transporting SCSI commands, data,and status messages over standard TCP/IPnetworks, it lets organizations take advantage ofexisting infrastructure and knowledge bases whileusing cost-effective, familiar components. It isalso standards based, which facilitates industryadoption and helps ensure interoperability. Andits performance can scale up along a commontechnology path—Ethernet.But the widespread use of iSCSI has alsocreated challenges as network usage has grownand evolved. For example, many organizationsare running an increasing number of rich-contentand data-intensive applications or virtualizedenvironments on their servers, which cansignificantly increase the iSCSI-based networktraffic in a data center. To help keep up with thisgrowth, many IT departments have begun movingfrom Gigabit Ethernet (GbE) network controllersto 10 Gigabit Ethernet (10GbE) controllers. Underheavy workloads, however, traditional controllerscan consume significant amounts of serverprocessing power, reducing the amount availablefor critical applications. In addition, the traditionalapproach typically requires using separate networkcontrollers to handle different types of tasks (suchas high-speed networking, storage, and clustering),which leads to a burgeoning number of devices,which drives up cost and complexity—and limitsthe ability of IT departments to take advantage of10GbE controllers.The Broadcom NetXtreme II family of 10GbEconverged network interface controllers (C-NICs)is designed to address these problems. Available inDell PowerEdge servers as LAN on Motherboards(LOMs), mezzanine cards, and stand-up networkinterface cards (NICs), these controllers providethe speed and efficiency to support heavy networktraffic workloads—enabling IT departments tosimplify their systems and provide network,storage, and clustering capabilities over existingTCP/IP and Ethernet infrastructures, while alsohelping reduce power consumption and enablinghighly efficient use of processing resources.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 75

Storage optimization“As converged controllers,Broadcom NetXtreme IIC-NICs can handle multipletypes of networking taskssimultaneously.”Easing the network processing loadAs converged controllers, Broadcom NetXtreme IIC-NICs can handle multiple types of networkingtasks simultaneously. A traditional server istypically equipped with four GbE controllers,along with other storage adapters in some cases.Simultaneously running network, storage, andclustering traffic at high rates typically requiresmultiple adapters and a large number of CPUcycles. The Broadcom C-NIC approach, incontrast, enables administrators to use a single10GbE port for multiple traffic types. Alternatively,if administrators choose to run just one traffic type,they can do so without having to deploy customhardware and software, which helps simplifyIT deployment and management. In addition,NetXtreme II C-NICs available in Dell PowerEdgeservers can support Microsoft ® Windows ® , Linux ® ,VMware ® , and other platforms, helping maximizedeployment flexibility.To help reduce the burden of packet headerprocessing on the host server, NetXtreme II C-NICsprovide iSCSI host bus adapter (HBA) functionalitywith iSCSI Offload Engine (iSOE) technology (seeFigure 1). By offloading iSCSI header processingfrom host processors to HBAs, these controllerscan help optimize server processor utilizationwhile helping increase both performance andthroughput for file-oriented storage, block-orientedstorage, backups, database transactions, andtightly coupled distributed applications such ashigh-performance computing workloads. The iSOEtechnology is designed to free up host processorcores and memory resources and increase I/Osper second (IOPS)—including, in the Broadcom testenvironment detailed in the next section of thisarticle, enabling up to 400,000 IOPS at 10GbE linerates over a single Ethernet port while substantiallyreducing processor utilization.iSOE enables NetXtreme II C-NICs to effectivelyhandle both TCP/IP and iSCSI processing. Byoffloading the TCP/IP and iSCSI stacks, the controllerdoes not need to compete with upper-layersoftware such as e-mail or Web applications forCPU cycles: iSCSI performance is unaffected byapplication workload. The convergence of blockstorage and network processing over a standard TCPinfrastructure helps eliminate the need for a separatestorage adapter and additional cabling, whileproviding performance and reliability comparable toFibre Channel at a significantly reduced cost.Broadcomcontroller modelPortsBroadcompart numberDell partnumber Card type I/O bus Physical interface iSCSI featuresBCM5709 Two GbE BCM95709A0907G 430-3254 Standard PCIe PCIe 1.0 1000Base-T iSCSI bootBCM5709 Two GbE BCM95709A0916G 430-3263 Mezzanine PCIe 1.0 1000Base-T iSCSI boot and HBABCM5709 Two GbE BCM95709SA0908G 430-3310 Blade mezzanine PCIe 1.0 SerDes iSCSI boot and HBABCM5709 Two GbE BCM95709A0907G 430-3261 Standard PCIe PCIe 1.0 1000Base-T iSCSI boot and HBABCM5709 Four GbE BCM95709A0906G 430-0800 Standard PCIe PCIe 1.0 1000Base-T iSCSI boot and HBABCM57710 One 10GbE BCM957710A1022G 430-2834 Standard PCIe PCIe 1.0 10GBase-T —BCM57710 Two 10GbE BCM957710A1021G 430-2836 Mezzanine PCIe 1.0 10GBase-T iSCSI boot and HBABCM57711 Two 10GbE BCM957711A1123G 430-0674 Blade mezzanine PCIe 2.0 10GBase-KX4 iSCSI boot and HBABCM57711 Two 10GbE BCM957711A1113G 430-0710 Standard PCIe PCIe 2.0 SFP+ iSCSI boot and HBAFigure 1. iSOE-enabled Broadcom NetXtreme II C-NICs available in Dell PowerEdge servers76 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

IOPS target configurationThroughput target configurationTarget 1 (1 Gbps)Target 1 (1 Gbps)Target 9 (1 Gbps)Target 2 (1 Gbps)Target 2 (1 Gbps)Target 10 (1 Gbps)Target 3 (1 Gbps)Target 3 (1 Gbps)Target 11 (1 Gbps)Target 4 (1 Gbps)Target 4 (1 Gbps)Target 12 (10 Gbps)Target 5 (1 Gbps)Target 5 (1 Gbps)Target 6 (1 Gbps)Target 7 (1 Gbps)Target 8 (10 Gbps)Target 6 (1 Gbps)Target 7 (1 Gbps)Target 8 (10 Gbps)Figure 2. High-levelconfiguration used in the testenvironment10GbE switchPortunder testSystem under test(Dell PowerEdge R710)Easing management and reducing costsare ongoing goals in the data center—andto that end, the NetXtreme II family helps tosimplify administration of controllers across theenvironment through the Broadcom AdvancedControl Suite (BACS) 3 management application,which provides a single platform for network andiSCSI HBA I/O management. BACS 3 includesat-a-glance status reports of LAN adapters andcontrollers in a system, network testing to confirmconnectivity to a remote station, and detailedperformance statistics on adapters and controllers.Evaluating the converged approachTo help IT departments understand theenhancements and benefits of the convergedapproach, in October 2009 Broadcom’sperformance laboratories ran a series oftests to analyze the power consumption,processor utilization, throughput, and processoreffectiveness (IOPS per CPU cycle) of aBroadcom NetXtreme II C-NIC and a third-partyNIC with an iSCSI software initiator.Figure 2 shows the test environment, which wasdesigned to evaluate the maximum performancevalue organizations could expect in this type ofdeployment. The system under test was a DellPowerEdge R710 server with two quad-coreIntel ® Xeon ® X5570 processors at 2.93 GHz, 12 GBof RAM, and the Microsoft Windows Server ® 2008Enterprise Edition OS. The setup included oneNetXtreme II BCM57711 dual-port 10GbE C-NIC withiSCSI HBA functionality enabled, along with a thirdpartyNIC using an iSCSI software initiator. The testsused 8 targets for IOPS measurement and 12 targetsfor throughput measurement; the target systemswere PowerEdge R710 servers with two quad-coreIntel Xeon X5570 processors at 2.93 GHz, 2 GB ofRAM, the Red Hat ® Enterprise Linux 5.2 OS, and iSCSIEnterprise Target (IET) 0.4.16 software.The tests utilized the Iometer I/O subsystemmeasurement and characterization tool, which isdesigned to evaluate the performance of storageand networked applications, perform stress tests onstorage and network devices, and predict storageand networked application performance. Thetest environment used version 2006.07.27 of thistool, configured for 128 outstanding I/Os. Powerconsumption was measured as the average of threeone-second samples taken at the power inlet.As these tests showed, using NetXtreme IIC-NICs as iSCSI HBAs can provide a varietyof advantages in data center environments,including helping to reduce power consumptionand increase processing efficiency. In the testenvironment, the NetXtreme II C-NIC in iSCSI HBAReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 77

Storage optimization240Third-party NIC with iSCSI software initiatorBroadcom NetXtreme II C-NIC with iSOEPower consumption (W)230220210200190180170160Typical I/O size for Web, paging, e-mail,and database workloadsTypical I/O size for OS, file copy, file server,and database workloadsTypical I/O sizefor streamingmedia, videorendering, andIP television(IPTV) workloads150140512 1,024 2,048 4,096 8,192 16,384 32,768 65,536 131,072 262,144 524,288 1,048,576I/O size (bytes)Figure 3. Power consumption at different I/O sizes for the third-party NIC and the Broadcom NetXtreme II C-NICmode provided a system-levelsavings of approximately 60 Wper port compared with thethird-party NIC (see Figure 3)—a difference that can lead todramatic cost savings. Forexample, consider a datacenter with 2,500 serverscontaining four 10GbE portseach. Assuming the same loadcharacteristics over the sameperiod as the test environment,using this NetXtreme II C-NICin place of the third-partyNIC could reduce powerconsumption by 600 kW at theport level alone; when factoringin an example Power UsageEffectiveness (PUE) value of 1.8and assuming matching savingsin power consumption acrossthe supporting infrastructure,the total reduction would reach1,080 kW. At an energy cost ofUS$0.088/kWh, then, the costsavings over the course of a yearwould amount to US$832,550.IT departments could also takeadvantage of these enormouspower savings to help themrightsize the data center’ssupporting infrastructure,helping to reduce total costof ownership and improvePUE efficiency.In addition to reducingpower consumption, theNetXtreme II C-NIC reducedprocessor utilization byapproximately 30 percent atlarge I/O sizes (2–64 KB) anddelivered three times the IOPSper CPU cycle of the thirdpartyNIC, enabling highlyefficient processing of storageworkloads (see Figure 4). Thisefficient operation in turncan provide increased IOPSperformance to end users forstorage workloads while alsofreeing up processing power tosupport additional applicationsProcessor effectiveness (IOPS per CPU cycle)14,00012,00010,0008,0006,0004,0002,0000Third-party NIC with iSCSI software initiatorBroadcom NetXtreme II C-NIC with iSOETypical I/O size for Web, paging,e-mail, and database workloads512 1,024 2,048 4,096 8,192 16,384 32,768 65,536 131,072 262,144 524,288 1,048,576I/O size (bytes)Typical I/O size for OS, file copy,file server, and database workloadsTypical I/O sizefor streamingmedia, videorendering, andIPTV workloadsFigure 4. Processor effectiveness at different I/O sizes for the third-party NIC and the Broadcom NetXtreme II C-NIC78 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

US$7,000,000US$6,000,000Space savingsPersonnel savingsSwitch purchase savingsServer purchase savingsUS$5,000,000Total savingsUS$4,000,000US$3,000,000US$2,000,000US$1,000,000US$01,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000Number of servers deployedFigure 5. Example onetime savings possible from proper planning and efficient processor utilization with BroadcomNetXtreme II C-NICsor increased usage of high-performanceapplications—all without increasing the total costof operations.This efficient operation can also help ITdepartments during installation, allowing them toplan for increased load capacity. In a data centerwith 5,000 servers, for example, the onetimesavings from proper planning and improvedutilization could reach US$5 million (see Figure 5). 1traffic onto a single network or use a dedicatednetwork for each. This approach holds thepromise of accelerating iSCSI networkingadoption and helping IT departmentspursue a variety of critical initiatives, fromenvironmentally friendly computing tovirtualization and consolidation—and theconstant need to keep costs down whilemaximizing data center efficiency.Reducing costs through converged networkingConverging multiple functions over one wireusing Dell PowerEdge servers with BroadcomNetXtreme II C-NICs can help simplify networking,increase processor effectiveness, and reduce totalcost of ownership. In addition, by enabling ITdepartments to take advantage of existing Ethernetinfrastructure and avoid the need for stand-aloneHBA cards, these converged controllers help lowerthe cost of iSCSI acquisition, deployment, andmanagement—avoiding the need to maintain aseparate storage infrastructure and the need forspecialized training.The Broadcom NetXtreme II family canalso enable organizations to take advantage of10GbE iSCSI networking while providing theflexibility to either converge data and storageLearn moreDhiraj Sehgal is a senior product linemanager for Ethernet controllers atBroadcom.Abhijit Aswath is a senior product linemanager for Ethernet controller software atBroadcom.Srinivas Thodati is a senior productmarketing manager for PowerEdge M-Seriesservers at Dell.Broadcom Ethernet controllers:broadcom.com/products/Ethernet-ControllersDell PowerEdge servers:dell.com/poweredge1 Based on a 30 percent reduction in server count from an average 30 percent reduction in processor utilization. Calculations assume an average servercost of US$5,000, a server lifetime of 3 years, 1 switch for every 2.5 servers, an average switch cost of US$1,500, 12 servers per rack, 10 square feetof rack floor space at a commercial cost of US$1,000 per square foot, 1 administrator for every 250 servers, and an average administrator salary ofUS$60,000.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 79

Storage optimizationAddressing datagrowth challenges withSymantec deduplicationsoftwareBy Charles ButlerCombining the Symantec Backup Exec 2010Deduplication Option with the Dell PowerVault DL2100 –Powered by Symantec Backup Exec can help dramaticallyreduce bandwidth and storage requirements for backupinfrastructures while providing versatile, enterprise-classdata protection.Although data growth is not new, thepace of that growth has becomemore rapid, the locations of data moredispersed, and the links between datasets more complex than ever—leaving enterprisesconstantly seeking ways to overcome their dataprotection challenges. Data deduplication offersthe opportunity to dramatically reduce the amountof bandwidth and storage required for backupsand to centralize backup data in support of disasterrecovery strategies. Although this technology hasexisted for several years, many organizations haveyet to take advantage of the enhanced efficienciesit can provide.Dell and Symantec have worked togetherto address the explosion of enterprise data bycreating the Dell PowerVault DL2100 – Poweredby Symantec Backup Exec. The combination ofthis integrated backup-to-disk solution and theBackup Exec 2010 Deduplication Option enablesorganizations to flexibly deduplicate data acrosstheir backup environments through either clientdeduplication (also called source deduplication,which occurs at the remote client servers usingBackup Exec agents) or media server deduplication(also called target deduplication, which occursin-line when a Backup Exec media server writesdata to disk).To help simplify deployment and scalability,the Backup Exec Deduplication Option doesnot use capacity-based licensing; instead, it isdelivered through a single license key that enablesall deduplication functions within Backup Exec,including client deduplication, media serverdeduplication, and appliance deduplicationthrough integration with Symantec OpenStorage–enabled intelligent disk devices (see the “Intelligentbackups through Symantec OpenStorage” sidebarin this article). Incorporating deduplication intothe Backup Exec platform can provide a variety ofadvantages in enterprise backup environments,including helping organizations to dramaticallyreduce bandwidth and storage capacityrequirements, centralize and manage backupoperations, and reduce rotations and use of tapefor disaster recovery.80 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

BackupExecagentBackupExecagentFigure 1. Client deduplication performed on thesource servers is designed to reduce the amountof data sent over the network, helping optimizeboth bandwidth and storage usageMicrosoftExchangeBackupExecagentMicrosoftSQL ServerBackupExecagentBackupExecagentMicrosoftOfficeSharePointBackupExecagentMicrosoftActiveDirectoryDell PowerVault DL2100 –Powered by Symantec BackupExec with Backup ExecDeduplication OptionMicrosoftHyper-VOracleDatabaseDeduplication andcompressionDeduplication and compressiontechnologies can both helpreduce the size of a backup,but the two approaches differin key ways. Compressioncan be applied to files,directories, or even volumes,but typically lacks awarenessof the underlying data andtherefore cannot recognize thatidentical files exist in differentdirectories. In addition, a typicalcompression algorithm—unlikethe deduplication process—canneither recognize changed datanor capture these unique blocksat a subfile level.When administrators backup data on a remote serverand choose to compressthat data before backing itup, the compression wouldgenerally result in reducedsizes for files sent to the localbackup application or acrossthe network to a centralizedbackup application. Whenusing Symantec Backup Execdeduplication, in contrast, fileson the remote server mightbe recognized as identical tofiles already stored, beforecompression occurs—in which case only a very smallamount of data would need tobe transmitted and stored. Othersystems sending data to thesame Backup Exec media serverwould also be aware of similardata that may have already beentransmitted and stored, helpingreduce backup sizes acrossthe environment.Bandwidth andstorage optimizationAdministrators can deploySymantec Backup Exec 2010deduplication to help optimizeboth bandwidth and storageusing client deduplication, or tohelp optimize only storage usingmedia server deduplication. Thedegree of storage optimizationcan depend on factors such asthe type of data, data changerate, and retention period.The Dell PowerVault DL2100 – Powered by SymantecBackup Exec provides an integrated backup-to-disksolution that supports flexible deduplicationReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 81

Storage optimizationSetting up Backup Exec 2010 deduplicationThis Dell TechCenter video walks through how to set up and configure adeduplication storage folder in Backup Exec 2010, as well as how to configurebackup jobs to utilize deduplication.delltechcenter.com/video/8866943/backup+exec+2010+set+up+de-duplicationBackup Exec client deduplication is supportedon Microsoft ® Windows ® OS–based systems,and requires a Backup Exec client agent on thesystem to be protected (see Figure 1). This agentenables the data to be deduplicated at this sourceclient before it is transmitted to the backup target,helping significantly reduce the amount of datathat must be sent over the network. In addition,Backup Exec agents are designed to leveragethe deduplication of data across all systemsunder protection in the environment, helpingachieve higher levels of deduplication than wouldotherwise typically be possible.The bandwidth reductions possible whenusing Backup Exec client deduplication make thisapproach well suited for protecting distributedservers or virtual machines with bandwidth (andI/O) constraints. Traditional full backups send allof the data across the network to the backupapplication, which then directs the data to amedia source where it can be compressed ordeduplicated; traditional incremental backups,similarly, transmit entire files during the backupprocess even though only a small portion ofa given file may have changed. Deduplicatingthis data at the source enables Backup Exec toperform full backups that transmit only data thatdoes not already exist on the backup target, andto perform incremental backups that send onlychanged segments of each modified file ratherthan the entire file.In many data centers, bandwidth constraintsmay not be a concern. In this type of environment,administrators may prefer to use Backup Execmedia server deduplication, which performsthe deduplication at the storage target after thedata has already been sent over the network(see Figure 2). This approach is typically easierto incorporate into existing backup architecturesthan client deduplication. 1Figure 2. Media serverdeduplication performedon the Dell PowerVaultDL2100 helps optimizestorage while simplifyingdeployment in existingbackup infrastructuresMicrosoftExchangeMicrosoftOfficeSharePointMicrosoftSQL ServerMicrosoftActiveDirectoryDell PowerVault DL2100 –Powered by Symantec BackupExec with Backup ExecDeduplication OptionMicrosoftHyper-VOracleDatabase1 For more information on types of deduplication and the advantages and disadvantages of different approaches, see “Demystifying Deduplication,”by Joe Colucci and Kay Benaroch, in Dell Power Solutions, 2010 Issue 1, dell.com/downloads/global/power/ps1q10-20100235-colucci.pdf.82 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Versatile platform and application protectionThe Backup Exec 2010 Deduplication Option supportsdata protection for a variety of platforms and applications.Administrators can enable either client or media serverdeduplication to help protect Microsoft platforms andapplications such as Hyper-V , Exchange, SQL Server ® ,SharePoint ® , and Active Directory ® software; VMware ® virtualmachines (with an agent installed in the guest OS); SymantecEnterprise Vault , SAP ® , IBM ® Lotus ® Domino, and IBM DB2software; and Oracle ® databases on Microsoft Windowsoperating systems. It additionally supports media serverdeduplication for VMware virtual machines using vStorageapplication programming interfaces (APIs); Linux ® , UNIX ® , andNovell ® NetWare ® operating systems; and Oracle databases onLinux operating systems.Efficient backups and data protectionFor IT environments where bandwidth is at a premium,combining the Dell PowerVault DL2100 – Powered bySymantec Backup Exec with client deduplication through theBackup Exec 2010 Deduplication Option enables bandwidthefficient,storage-optimized data protection for remotesystems and virtualized environments. For environmentswhere bandwidth is not a concern, Backup Exec media serverdeduplication enables efficient use of storage resources andcan help simplify deployment in existing backup infrastructures.Although specific results can vary depending on type of data,change rate, retention period, and other factors, SymantecBackup Exec 2010 deduplication technology in theDell PowerVault DL2100 can help organizations create a flexiblebackup infrastructure to help maximize management andstorage cost efficiencies.Learn moreCharles Butler is a technical director in theInformation Management Group at Symantec.Symantec Backup Exec:backupexec.comDell PowerVault DL2100:dell.com/dl2100Dell and Symantec:dell.com/symantecsymantec.com/dellIntelligent backupsthrough SymantecOpenStorageThe Symantec OpenStorage initiative is adisk-based innovation that allows SymantecBackup Exec to take advantage of the advancesof intelligent disk storage devices—includingstorage reduction, backup image duplication,synthetic backups, replication, and energyefficiency. The tight integration with thirdpartyintelligent disk storage devices providesenhanced management of backup images andadditional functionality while helping avoid thelimitations of tape emulation.Intelligent disk storage vendors writeplug-ins for the OpenStorage applicationprogramming interface (API) that integratewith Backup Exec. These plug-ins provideBackup Exec with visibility into the propertiesand capabilities of the storage devices as wellas control over the backup images stored onthem. Backup Exec can then treat them as diskdevices rather than tape devices, as in the caseof virtual tape libraries.Using OpenStorage plug-ins and theOpenStorage API, Backup Exec can controlwhen backup images are created, duplicated,and deleted, while the intelligent diskstorage devices control how the images areactually stored in and replicated betweendevices. In addition, intelligent disk storagemanufacturers in the OpenStorage programcan add value to an overall solution throughspecialized innovations such as backup imagededuplication, wide area network (WAN)–optimized backup image replication for disasterrecovery, power management, and more.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 83

Storage optimizationEnd-to-end, snapshot-awaredata protection withCommVault SnapProtectBy Darin CampIn conjunction with Dell EqualLogic PS Series snapshotcapabilities and CommVault ® Simpana ® SnapProtect technology, the Dell PowerVault DL2100 – Poweredby CommVault enables end-to-end data protectionthat supports fast, flexible recovery and aggressiveservice-level agreements.Today’s IT departments are facingexplosive data growth along withservice-level agreements (SLAs) for dataavailability and recovery that can oftenseem unobtainable. Because traditional tape-basedprotection typically captures data only from theprevious backup event, a recovery request maymean a gap of up to 24 hours from the mostrecent recovery point—often unacceptable forbusiness-critical data.Moving the primary recovery location of criticaldata to a readily available storage area network(SAN)–based snapshot can significantly mitigatethe risk of data loss. In conjunction with DellEqualLogic PS Series Internet SCSI (iSCSI) SANarrays and CommVault Simpana SnapProtecttechnology, the Dell PowerVault DL2100 – Poweredby CommVault can help organizations create anend-to-end environment that leverages nativeEqualLogic snapshots to enable fast, flexible dataprotection and recovery. As an integrated, fullfeaturedbackup appliance, the PowerVault DL2100is designed to significantly simplify storage andbackup administration and help create an advanced,highly scalable storage solution. By incorporatingEqualLogic snapshot copies, the PowerVault DL2100with SnapProtect, and multiple copy tiers into theenvironment as part of a comprehensive recoverystrategy, IT administrators can meet the challengesof data growth while supporting aggressive SLAs anddata protection requirements for critical systems.Enabling end-to-end, snapshot-awaredata protectionThe separation of recovery-copy, protectioncopy,and archive-copy tiers provides theavailability and speed required for today’s criticalapplications, and helps reduce costs by usingthe appropriate storage types and technologiessuch as deduplication to meet long-termretention requirements (see Figure 1). Formingthe foundation of the local recovery-copy tier,snapshots capture the disk layout at the exactpoint the snapshot is initiated, providing a point-intime(PIT) copy that can then serve as a referencefor the other tiers. These PIT copies can be easilymounted to the Dell PowerVault DL2100 fornightly backups—thus enabling IT staff to use thesame consistent snapshot for off-site recovery andretention copies on cost-effective disk and tape.84 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

With the SnapProtectEnabler (SPE) introduced inCommVault Simpana 8.0, SANsnapshot management controlsare coded directly into the dataprotection agents—deliveringcombined protection andrecovery capabilities within asingle operation from beginningto end. The SPE is the corepiece of intelligence creatingthe CommVault open snapshotmanagement framework.Enabling it is a simple checkbox option within the SimpanaIntelligent Data Agent (iDA) thateffectively takes the backupprocess and enhances it withthe ability to integrate nativeSAN array snapshots.The SPE integrates nativelyinto host systems, applications,and SAN arrays to blendthe speed and efficiency ofhardware snapshots with robustSimpana data access capabilities.Instead of scheduling multipleprocesses with differentsolutions or scripts, Simpanasoftware is designed to controlthe entire progression of datawithin a single protection policy(see Figure 2). This single-stepprocess enables administratorsto create, retain, and catalogthe data within snapshotswithout the need to compile ascript or take on unnecessarymanagement overhead.From a single protectionpolicy, administrators canleverage snapshots on theSAN by adding SnapProtectto the backup client andthen configuring it for theappropriate snapshot applicationprogramming interface (API).These snapshots are mounted,cataloged, checked for dataintegrity, and retained as ahighly available recovery copy(see Figure 3). By using thissame copy, the software movesthe data to the appropriateprotection copy or archive copybased on the original protectionpolicy. Data accessibility remainsthe same regardless of the datalocation. The simple browseand recovery from a traditionalbackup extends to the snapshotcopies, providing IT teams withRecovery-copytier• Provides a local copyfor rapid recovery• Can include snapshotsand replicas• Can be mounted fornightly backups• Retains data for aspecified number ofhours or daysFigure 1. Data management copy tiers optimized for different scenariosa simplified, cost-effective wayto incorporate snapshots intoa tiered strategy to help meetrecoverability requirements.Implementing a protectioncopytier with deduplicationon disk and tape providesa long-term, versionedretention strategy to help meetorganizational and regulatorycompliance requirements whileData deduplicationData encryptionContent indexingProtection-copytier• Provides a restorecopy (full system, store,and object) and/ordisaster recovery copy(remote site)• Includes tieredretention• Retains data for aspecified number ofdays or weeksArchive-copytier• Provides a retentioncopy (shared objects)• Supports secure accessfor recalls or indexedsearch• Includes tieredretention (nearline,offline, and vaults)• Retains data for aspecified number ofdays, months, or yearsRecovery copyD:\Production storeD’D’D’’D’’’Data center requirement Retention requirement Advantages• 4-hour recovery pointobjective (RPO)• Less than 4-hour recoverytime objective (RTO)• Revert operation• 48 hours of 4-hourchange increments• Supports rapid recovery fromdata corruption or loss• Enables fast off-host protectionpolicies• Can be easily mounted andthen recoveredProtection copyOff-hostD’’’AuxiliarycopyDeduplicateddisk libraryTapelibrary• 24-hour RPO• 24-hour RTO• Daily off-site backups• On-site deduplication• 4–6 weeks of dailychange increments• 3 weeks of daily off-sitebackups• Supports recovery fromhardware failure• Uses cost-efficient storage• Uses deduplication to enablemulti-version copies• Helps simplify disaster recoveryFigure 2. Recovery-copy,protection-copy, andarchive-copy tiers as part of acomprehensive data protectionand recovery strategyArchive copyAuxiliarycopyDeduplicatedtape library• Archive RPO (weekly full,monthly full)• Multiple-business-day RTO• Deduplicated consolidationon tape• 1–2 years of weeklyor monthly full databackups• Helps meet compliancerequirements• Enables cost-effective legal holdretention• Is designed for portability,cost-effectiveness, and energyefficiencyReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 85

Storage optimizationSimplifyingarray-basedsnapshotsArray-based snapshotscan have significantadvantages in enterpriseIT environments, but canalso increase operationalcomplexity. This videofrom the CommVaultSimpana Whiteboard Seriesexplores how SnapProtecttechnology helps overcomethese challenges byenabling administratorsto control array-basedsnapshots through a singlemanagement console.download.commvault.com/unsecure/media/video/whiteboards/snapprotect/“Layering deduplication withcost-effective storage canhelp drastically reduce costsfor long-term retention.”also helping protect criticalapplications from a productiondisk failure. Recoveries from thistier typically take longer thanSAN-volume revert operations,but the disaster recovery copiesand off-site replication canenable additional data availabilityoptions for the longer SLAs at amuch lower cost compared witha recovery-copy tier.Further extending the dataprotection methodologies toan archive-copy tier provides along-term, removable data copyto support retention policiesand regulatory compliance.Layering deduplication withcost-effective storage can helpdrastically reduce costs for longtermretention. Allowing a singleprotection policy to initiate theprogression of data through therecovery-copy, protection-copy,and archive-copy tiers withoutsacrificing data availabilityhelps optimize current andfuture storage hardwareimplementations.SnapProtect blends theintelligence of Simpanaapplication iDAs with nativesnapshot capabilities tocreate application-consistentPIT data copies. Applicationawareness enables SnapProtectto incorporate a PIT recoverycopy with the cataloging of astandard backup routine from ascript-free snapshot protectionoperation. Recovering data to aPIT is a single operation througha volume revert operation, arecovery from deduplicateddata on disk or tape, or a blendof the two, depending on thespecific recovery point. Theprocess of sequencing andscripting application quiescence,initiating the snapshot, mountingthe snapshot, and runningprotection and/or recoveryprocedures is managed throughthe CommVault CommCell ®Figure 3. Hardwaresnapshot processusing Dell EqualLogicPS Series storagewith the DellPowerVault DL2100 –Powered byCommVaultCommVaultSimpana SPEApplicationsDatabasesMessagingand documentsFile systemsOnline copy Nearline copy(snapshots on (selective copyAproductionBOffline copyon deduplicatedC(offline media)storage)disk)ActivedataSnapshotdataVolumes–6 hours–12 hours–18 hoursDell EqualLogic PS SeriesiSCSI SAN array (production)ADell PowerVaultDL2100 – Powered byCommVault (proxy)BABCSnapshots are initiated basedon the defined schedule, afterwhich the Dell PowerVault DL2100mounts and indexes the snapshotsas a protection copyThe protection policy ages theonline snapshot copies and createssecondary copies for retention onthe protection-copy storage tierData can be recovered across therange of copies, including one-clickrecovery of entire volumes orapplications as well as individualobjects from online snapshots orbackup mediaDeduplicateddisk libraryCTapelibrary86 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Snap protectionclient (SPC) groupMicrosoft Exchangeclient groupOracle Databaseclient groupSingle console for mixedapplications, operatingsystems, and hardware; localand remote sites; copies; reportingand auditing; scheduling andcoordination; tiered media;deduplication; and encryptionSPC(SPE)SPC(SPE)SPC(SPE)SPC(SPE)SPC(SPE)Volume Volume VolumeDell EqualLogic PS SeriesiSCSI SAN arraySPC(SPE)SPC(SPE)Storage policy (synchronous copies)Protection or archive copies(standard media or deduplicated store)Dell PowerVault DL2100 –Powered by CommVault (proxy)Figure 4. Comprehensivemanagement through a singleCommVault Simpana consoleinterface without requiringscripting. Simpana can access thecopy or copies (recovery copy,protection copy, or mixed access)necessary in a one-pass routinewith minimal IT staff intervention.Spotlighting MicrosoftWindows applicationsCommVault Simpanasupports software includingMicrosoft ® SQL Server ® ,Microsoft Exchange, Oracle ® ,and SAP ® applications as wellas file systems on MicrosoftWindows ® , Linux ® , and UNIX ®operating systems. In particular,Microsoft Volume ShadowCopy Service (VSS) integrationwith the CommVault snapshotmanagement framework helpsensure tight application controlfor a snapshot routine onWindows operating systems.Taking advantage of VSS createsa simplified adoption pathfor organizations seeking toeliminate scripted snapshotoperations or implement highlyavailable recovery copies for thefirst time. By simply enablingSnapProtect, administrators canenable applications to becomesnapshot aware with the backendstorage array.Supporting comprehensivedata protection and recoveryOrganizations seeking tomeet stringent SLAs whileenhancing data protectionand application uptime nolonger need solutions frommultiple vendors, dedicatedhardware, and custom scriptingto make everything worktogether. With SnapProtect,the Dell PowerVault DL2100– Powered by CommVaultenables IT administrators toincorporate comprehensivesnapshot awareness intotraditional backup and recoveryroutines while still retainingthe advantages of applicationdata file consistency and logmanagement. LeveragingDell EqualLogic iSCSI SANarrays with the CommVaultopen snapshot managementframework provides a singleprotection policy that createshighly available local recoverycopies, protection copies foroff-site storage and disasterrecovery, and archive copiesfor long-term retention (seeFigure 4). The integrationwith the PowerVault DL2100– Powered by CommVaultprovides a simplified,integrated way to implementscript-free data management(including deduplication) andcontrol processes throughoutthe data life cycle—with theadded advantages of reducedbackup media costs, shortenedbackup windows, andreduced bandwidth replicationrequirements.Darin Camp is a seniortechnical alliancemanager at CommVault,and has spent the past12 years in the storagenetworking and dataprotection industries.Learn moreDell and CommVault:dell.com/commvaultDell PowerVault DL2100:dell.com/dl2100Dell EqualLogic PS Series:dell.com/psseriesdell.com/equallogicReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 87

VirtualizationOvercoming data protectionchallenges in virtualizedenvironmentsBy Scott HeroldRelying on traditional backup approaches in virtualizedenvironments can be complex and time-consuming. Byadopting image-based data protection technologies, ITstaff can create a simplified, streamlined, reliable system forefficiently protecting and recovering virtual machines.VizioncorevCommunityJoin the VizioncorevCommunity for the lateston protecting and managingvirtualized environments,including blogs, discussionforums, videos, and more.vcommunity.vizioncore.comWhen virtualization was firstmoving into the mainstream, ITgroups were largely concernedwith “Step 1” problems—determining how to plan and deploy virtualizedservers to consolidate physical infrastructure andreduce costs. These problems were generallystraightforward to solve, and many organizationswere able to achieve the types of consolidationbenefits that they expected.Now, however, these IT groups are deepinto the challenges of “Step 2”—operating thoseenvironments efficiently. And the problems thathave come up are not so simple to solve. Virtualmachine (VM) sprawl, for example, took manyIT teams by surprise. Setting up a new VM issignificantly faster, easier, and more cost-effectivethan requisitioning an entire physical system,which is the analogous process in a traditionalnon-virtualized environment. Application ownerscan therefore obtain the necessary development,quality assurance, and production applicationservers relatively easily, which can quickly leadto an explosion of VMs. In a data center that hasconsolidated 50 physical servers to 4 virtualizedservers, IT staff might expect a correspondingreduction in their administrative burden. Butbecause the operating systems themselves don’tgo away, the workload actually continues to growas these additional VMs are created. Six monthsin, this environment might have 100 VMs—or whatamounts to double the workload.Another Step 2 problem—and the one thatis the focus of this article—is data protection. It’sno surprise that many organizations still protectdata in VMs as if they were working with physicalsystems, by deploying backup agents into theVMs and then using those agents to back upand recover files while often doing nothing toprotect the VM image itself. But by instead takingadvantage of image-based backup technologies,these organizations can overcome the numerousproblems with these traditional backup approachesand create a simplified, streamlined, reliableapproach to data protection in virtualizedenvironments based on Dell systems.Challenges of traditional backupsTraditional backup approaches are less thanideal in virtualized environments. Performing acomplete system recovery using file agents in adisaster recovery scenario, for example, can be88 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

difficult and extremely timeconsuming.In addition, becauseof the shared resource modelof virtualization, schedulingbackup jobs can be tediousor, in some cases, impossible.Backup administrators mustidentify the servers hosting eachVM—which may change in realtime in environments using livemigration technologies—andensure that only a certainnumber of backup jobs run atone time on a given server. Whenbackup agents are individuallydeployed in every guest OS,the backup job is consumingunderlying system resources theentire time it is running—slowingdown not only the VM beingprotected, but also other VMson the same server. Runningsimultaneous backup jobs ondifferent VMs on the same serveris generally out of the question.Then there is the questionof how backup data is movedfrom the virtualized serverto storage. In a physicalenvironment, a separate backupserver is typically attached toeach client to move that data.This approach can still workfor virtualized servers—butall backup data must be sentover the LAN, which candisrupt use of that network forother users. To combat this,some organizations may usea separate backup network,which increases the cost andcomplexity of the infrastructure.Another option has been to useVMware ® Consolidated Backup(VCB), but this feature requiresa storage area network andhas been superseded by theVMware vStorage applicationprogramming interfaces (APIs)introduced with the VMwarevSphere release.vStorage offers an enhancedway for backup software tocapture VM data, by workingwith the VM image througha published VMware API(see Figure 1). Image-basedbackup software, such asthe Vizioncore ® vRanger Pro Data Protection Platform(DPP) solution for image-basedbackup and recovery, can offermajor advantages in virtualizedenvironments. It can acceleratethe process of capturing data,helping reduce the impact onthe system and the organizationas a whole. It can fill a criticalgap by helping protect thefull VM image as well as theindividual files within that image.It can help speed recovery andincrease recovery reliability forindividual files as well as forthe entire image. And it canhelp increase data transmissionspeed—it’s generally far fasterto transmit the whole imageinstead of the many individualfiles that comprise the image.Advantages of image-baseddata protectionWhen designed well, imagebasedbackup can providea more reliable method forhandling data than traditionalmethods. What IT teamsworry about is the integrity ofthe backup data copies heldin the archive, and creatingthose copies depends on acontinuous, uninterruptedwrite of the backup data copy.Another concern is the integrityof the application data captured,Putting image-baseddata protection to workImage-based data protection can provide significantadvantages over traditional backups in a wide range oftypical IT scenarios.Rapid, efficient backup and recovery of large VMimages, such as those that exceed 10 GB in sizeSupport for local high availability of VMs bypreserving them and then restarting them on alternatevirtualized servers in the environmentDisaster recovery through the transmission of VMimages to an off-site recovery location, in which caserecovering both the data and applications occursin the single step of launching the replica VMs ondifferent serversRapid recovery of entire servers, especially thosehosting business-critical applications such as customerrelationship management, payroll, and accountingsystems, in which case the ease of recovery—withoutrequiring separate steps for operating systems andapplications—can make image-based approachesinvaluableRollback of database systems to a point in timejust before drive resizing and other optimizationtasks—just in caseProtection for end users and customers against diskproblems, data corruption, and the troubleshootingsteps required for repairs, replacements, and bringingstorage environments back to operational readinessRapid recovery from human errors such as storageprovisioning and reallocation by mistakenly deletinglogical units (LUNs) that contain critical data—a situationthat can occur frequently when virtualized servers areunderprovisioned and inexperienced administratorsattempt to find space for backups, or when the steps ofan infrastructure upgrade or cutover are performed outof sequenceReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved. dell.com/powersolutions | 2010 Issue 02 89

VirtualizationFigure 1. Enhanceddata capture,transmission, andrecovery processesusing image-based dataprotection in virtualizedenvironmentsSystem state, OS and application configurations,file system, and application data filesVirtualPhysicalData captureImagefilesData recoveryEnvironmentWhole systemSingle fileSingle objectRecovery locationLocal and remote locationsPhysical and virtual systemsPhysical-to-virtualand virtual-to-physicaltransformationswhich must be consistent at a given point in timeto be recoverable and usable from that point.Both problems can be alleviated with a welldesignedimage-based data protection system.First of all, this type of system can be extremelyfast—incurring minimal overhead by bypassingthe file system to directly read a disk, in additionto skipping empty blocks in the image. Thisapproach also creates a smaller backup copy—ina single file—than would be created by thetraditional method of creating a backup copy ofthe thousands or millions of individual files thatmake up the image. Faster means that there is lesstime for something to go wrong, while smallermeans that the backup copy can be transmittedmore quickly and stored using fewer blocks than itcould otherwise—again helping reduce the risk ofsomething going wrong.In terms of application data consistency,because capturing an image that includes all of anapplication’s data from a single point in time is thedefinition of an image-based backup, this approachhelps significantly reduce the odds of backing updata at inconsistent points. Even in the unlikelyevent that the data is not consistent, administratorscould still recover and restart the application byrolling back to the nearest previous consistencypoint, which is captured in the backup image.When asked to comment on image-baseddata protection, Lauren Whitehouse, a senioranalyst from the Enterprise Strategy Group(ESG), agreed that this technology is theone most likely to help improve backup andrecovery in virtualized environments. CitingESG’s 2010 IT spending survey, in which over500 respondents reported on their budgetsfor investing in virtualization technologythis year, she said, “ESG research found thatimplementing data protection processes forserver virtualization environments is a big painpoint for many end-user respondents. Theseorganizations have committed to fixing theissue this year with increased investments insolutions that improve backup and recovery ofvirtual machines, improving disaster recoveryprocesses, and improving application backupand recovery.” She continued, “For x86 servervirtualization environments, rapid image-levelbackup with flexible image- or item-level restorecan address all of these challenges. It canprovide nondisruptive and optimized backup,enable efficient disaster recovery strategies, andfacilitate improvements in application-specificbackup and recovery.” (For more examples of thebenefits of image-based data protection in realworldscenarios, see the “Putting image-baseddata protection to work” sidebar in this article.)Streamlined data protectionfor virtualized environmentsVizioncore has introduced its Backup 2.0 campaignto help organizations understand how imagebaseddata protection can simplify their operations.The sooner that IT teams begin to adopt imagebasedmethods for VM backups, the sooner theycan overcome the Step 2 challenges of protectingtheir virtualized Dell servers—a critical step towardgaining the full benefits of virtualization.Learn moreScott Herold leads product design andarchitecture for Vizioncore and has more thana decade of industry experience in OS, network,security, and storage design.Dell and Vizioncore:dell.com/vizioncoreVizioncore Backup 2.0Resource Center:vizioncore.com/backup2090 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

VirtualizationSimplifying networksin virtualized environmentswith Intel technologyBy Srinivas Thodati and Brian JohnsonAdvances in Intel ® 10 Gigabit Ethernet technologyand VMware ® vSphere virtualization can help create aflexible, simplified, highly efficient networking environmentwithout compromising areas such as security andtraffic segmentation.The success of virtualization, whilesubstantial, has been limited by thecomplexity that has arisen fromnetworking with Gigabit Ethernet (GbE)technology. Networking virtualized hosts with GbEis typically based on relatively large numbers ofdedicated physical connections, which are usedto segregate different types of traffic because ofnetworking infrastructure and bandwidth limitationsfrom 1 Gbps connectivity. Organizations pay forthis complexity in increased management andequipment costs, as well as increased power usage.As 10 Gigabit Ethernet (10GbE) technologyenters the mainstream, IT organizations areconsidering how they can use it to help simplifytheir virtualized environments. Consolidatingmultiple GbE traffic flows onto a reduced numberof 10GbE connections helps reduce complexityand cost. However, many IT administrators haveexpressed concerns that quality of service (QoS),security, or performance may suffer as a resultof this traffic convergence. These concerns haveprevented some organizations from taking fulladvantage of 10GbE technology.Cumulative experience among Dell,Intel, and VMware together with advancesin Dell PowerEdge servers, Intel 10GbEserver adapters, and VMware vSphere 4virtualization software can alleviate theseconcerns and help organizations realize furtherefficiency and cost savings. 1Transitioning to a virtualization mind-setThe common practice when deploying virtualizedhosts in a physical environment has been tosegregate network functions onto dedicatedGbE ports, adding additional ports as demandor bandwidth increases. These ports are ofteninstalled in pairs to provide network failover,doubling the number of ports required perhost. As a result, the number of network portshas a tendency to become bloated, leading toexcessive complexity and associated high costs ofhardware and power.1 To learn more about optimizing performance in virtualized environments with 10GbE connectivity, see “Intel Ethernet Server Adapters MaximizeThroughput Performance for iSCSI Connectivity,” by Sunil Ahluwalia and Gary Gumanow, in Dell Power Solutions, 2010 Issue 1, dell.com/downloads/global/power/ps1q10-20100322-Intel.pdf.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 91

VirtualizationEnabling near-native10 Gigabit EthernetperformanceDistributed Switch, using port 2. VM traffic canuse one or more VLANs, depending on the levelof separation needed between the VMs. In a nonfailedstate, the VMs should not share the serviceconsole, vMotion, or IP-based storage trafficVLAN IDs, and should use port 1 in a two-portconfiguration.VLANs for traffic isolationWhen VLANs are separated in this manner, logicalpartitioning isolates individual traffic flows. VMwarevSphere can control the effects of individual VMson the traffic flows of other VMs that share thesame physical connection. Demilitarized zones(DMZs) can be configured on different networkadapters to isolate internal traffic from external traffic.Realizing the performance benefits of 10 Gigabit Ethernet (10GbE) invirtualized environments requires minimizing software overhead. Severaltechnologies can enhance networking performance in these environments.Intel Virtual Machine Device Queues (Intel VMDq): Offloads data-packetsorting from a virtual switch onto a physical network adapter, helping toreduce processor overhead and enhance overall efficiency*Single Root I/O Virtualization (SR-IOV): Allows administrators to partitiona physical port into multiple virtual I/O ports, and use each as a singlededicated port directly assigned to a virtual machineIntel Virtualization Technology for Directed I/O (Intel VT-d): Enables an I/Oresource to be assigned to a specific virtual machine, giving its OS directaccess to that resource**Data Center Bridging (DCB): Provides enhanced traffic prioritization overa single interface and, in conjunction with 10GbE, can help cost-effectivelymeet latency requirements****To learn more about Intel VMDq, see “Intel Ethernet Server Adapters Maximize Throughput Performance foriSCSI Connectivity,” by Sunil Ahluwalia and Gary Gumanow, in Dell Power Solutions, 2010 Issue 1, dell.com/downloads/global/power/ps1q10-20100322-intel.pdf.**To learn more about Intel VT-d, see “Building the Efficient Enterprise with Dell, Intel, and VMware vSphere 4,”by Andrew Gilman, Mike Monthei, and Andrew I. Fields, in Dell Power Solutions, September 2009, dell.com/downloads/global/power/ps3q09-20090301-vmware.pdf.***To learn more about DCB, see “Mixing Gigabit Ethernet and 10 Gigabit Ethernet in a Dedicated SANInfrastructure,” by Tony Ansley, in Dell Power Solutions, September 2009, dell.com/downloads/global/power/ps3q09-20090416-ansley.pdf; and “10 Gigabit Ethernet: Unifying iSCSI and Fibre Channel in a SingleNetwork Fabric,” by Achmad Chadran, Gaurav Chawla, and Ujjwal Rajbhandari, in Dell Power Solutions,September 2009, dell.com/downloads/global/power/ps3q09-20090392-chadran.pdf.Administrative traffic and other back-end services arehandled by a separate networking stack managedby the VMkernel, providing further isolation from VMtraffic even on the same physical connection. 2Enterprise-level reliabilityMany IT organizations are concerned about theneed to ensure enterprise-level reliability for criticalapplications and workloads. To enhance reliability,administrators can install 10GbE ports in pairs tosupport a redundant configuration. If two 10GbEports are used, a best practice is to run VM trafficprimarily on port 1 and all other traffic on port 2.This design uses the bandwidth of both 10GbEports and can be configured for network failover.In the event of a hardware failure, the hostand management software should migrate allVMs off the host using vMotion to retainredundancy and help ensure reliability. Toenhance redundancy further, a second optionis to move to a configuration of four 10GbEports, with two primary ports and backup portson separate adapters.Meeting the bandwidth requirementsof converged data streamsAnother concern among many administrators isthat 10GbE connections can become saturatedwith one traffic type, negatively affecting othertraffic classes. In particular, live migration ofVMs using VMware vMotion can be bandwidthintensive. However, reaching saturation on 10GbEis unlikely, given that the converged traffic streamspreviously used only GbE connections.Using Dell PowerEdge servers with powerfulmulti-core processors and Intel 10GbE serveradapters can provide the overall bandwidth tosupport this aggregation. 10GbE adapters providethe bandwidth necessary for multiple traffic typesto coexist on a single port. In many cases, QoSrequirements can be met simply by the availabilityof large amounts of bandwidth. The presence ofsufficient bandwidth can also increase the speedof live VM migration using vMotion, removingpotential bottlenecks for enhanced performance.2 To learn more about using VLANs to protect data, see “vSphere 4.0Security Hardening Guide,” by VMware, Inc., April 2010, available atcommunities.vmware.com/docs/DOC-12306.94 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2.Copyright © 2010 Dell Inc. All rights reserved.

In addition to generally available bandwidth,traffic segmentation helps ensure dedicatedbandwidth for specific traffic types and QoS.Shared 10GbE server adapters are moreadvantageous than separate GbE connectionsfor this purpose. The 10GbE connections makemore efficient use of available bandwidth thanGbE connections for demanding applications andheavy-use portions of the workday. 3GbE limitations with trafficspike managementAlthough separating traffic flows onto discreteGbE connections is a viable means of providingdedicated bandwidth, doing so has distinctshortcomings. For example, allocating two GbEconnections to a VM traffic port group providesa potential of 2 Gbps of dedicated bandwidth.However, if additional bandwidth is needed forsporadic traffic spikes from that port group,additional server adapters must be added—assuming additional PCI slots are even available.Another shortcoming is that the bandwidthallocated in this example cannot be used by anyother traffic, so it simply goes to waste.10GbE and seamless managementof traffic spikesProviding bandwidth to the port group from ashared 10GbE server adapter allows additionalbandwidth to be allocated seamlessly for trafficspikes. Multiple port groups can share thebandwidth headroom provided by the serveradapter. The resource can be automatically anddynamically reallocated as various port groupsare accessed by their associated VMs. Dedicatedbandwidth is not necessary for any one port groupas long as the host network connection does notreach saturation.Deploying 10GbE to help reducecomplexity and costUsing Dell PowerEdge servers with Intel 10GbEserver adapters in VMware vSphere 4 virtualizedenvironments enables organizations to reducethe complexity and cost of their networkinfrastructure. These hardware and softwaretechnologies are designed to work togetherto ensure that security and performancerequirements can be met without the largenumbers of physical server connections requiredin legacy GbE networks. As a result, administratorscan replace methods that depend on physicalseparation with updated approaches that uselogical separation.The key for IT organizations is to resist tryingto mimic a traditional setup of physical serversin a virtualized server environment. Rather thansegregating a 10GbE port into multiple dedicatedchannels, they can take advantage of open10GbE connections and VMware virtualizationto enable flexible, dynamic bandwidth allocationand increased headroom while also dramaticallyreducing the number of physical server adaptersneeded for a given configuration—which, inturn, helps them to reduce capital and supportcosts. Ultimately, 10GbE connectivity enablesorganizations to obtain maximum value from theirvirtualized server environment.Learn moreSrinivas Thodati is a senior product marketingmanager at Dell, and has more than 16 years ofexperience in the IT industry.Brian Johnson is a product marketing engineerat Intel for 10 Gigabit Ethernet products andvirtualization technologies, and has more than12 years of experience in server product planningand marketing.Intel Ethernet server adapters:intel.com/go/ethernetintelethernet-dell.comIntel Virtualization Technology:intel.com/go/vtcVMware virtualization:vmware.com/virtualizationDell PowerEdge servers:dell.com/poweredge3 For more information on how 10GbE connectivity handles the demands of bandwidth-intensive applications, see “New Technologies Speed theMove to 10 Gigabit Ethernet Data Center Connectivity,” by Sunil Ahluwalia, in Dell Power Solutions, June 2009, dell.com/downloads/global/power/ps2q09-20090230-intel.pdf.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 95

Mobility solutionsProductivity atthe touch of a button:Dell Latitude ON | FLASHBy Roberto Ayala and Steve RokovThe Dell Latitude ON | FLASH module was designedfor on-the-move workers. By avoiding a full system boot,Latitude ON | FLASH is available within seconds, providingnear-instant access to the Internet, Microsoft ® Officedocuments, virtual remote desktops, and more.Dell Latitude ON | FLASH moduleBooting up a laptop just to check e-mail or access theInternet can be a slow process—and a definite drain onproductivity. On-the-move workers can reclaim that losttime by using Dell Latitude ON | FLASH, the latest memberof the Latitude ON family.Supported Latitude ON Ready laptopsinclude Dell Latitude E4200, LatitudeE4300, Latitude E4310, Latitude E6410(shown here), Latitude E6510, andDell Precision M4500 modelsAccess in secondsInstalled as a separate, alternative workspace designed to launch inunder 10 seconds, Latitude ON | FLASH offers a full-featured, secure,high-performance Web browsing experience together with Citrix ® ,VMware ® , and Remote Desktop Protocol (RDP) remote desktopsupport. Users can also check e-mail, access chat and voice-over-IP (VoIP) services, and read and edit Microsoft Office–compatibledocuments.Because it is designed to launch in seconds at the touch ofa button, Latitude ON | FLASH helps reduce the need to leavelaptops turned on or in sleep mode overnight—helping to lowerpower consumption and reduce boot problems with the MicrosoftWindows ® OS. And unlike a conventional mobile thin client, laptopsequipped with Latitude ON | FLASH transform into mobile hybridcomputing devices that provide the flexibility to work in a traditionalthick-client mode, a thin-client mode, or a combination of the two.Latitude ON | FLASH can be preinstalled on supported models atthe time of purchase or field installed as a module kit that includesthe necessary hardware, drivers, and application software. Afterconfiguring the software using a Windows utility, end users can96 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Figure 1. Dell Latitude ON | FLASH start screen, designed to launch within 10 secondslaunch Latitude ON | FLASH by pressing the Latitude ON button,entering their login information, and choosing an application fromthe start screen, without waiting for a full Windows boot (seeFigure 1). Users can switch easily between the Latitude ON | FLASHand Windows environments. And because the module utilizes thesystem power, processor, memory, LAN, and wireless networkingresources, users can continue to leverage their existing investments.Latitude ON | FLASH works with supported Latitude ON Readylaptops, easily identifiable by the separate Latitude ON button; thesemodels include Dell Latitude E4200 and Latitude E4300 laptops inaddition to the recently announced Latitude E4310, Latitude E6410,Latitude E6510, and Dell Precision M4500 models.Rock-solid reliabilityLaptops are in frequent motion and subject to regular bumps anddrops. Latitude ON | FLASH is designed to operate independently ofthe laptop’s primary OS and hard drive so that users can still get onlineeven following an OS or hard drive failure. Its alternative workspaceremains active even if the hard drive has been encrypted or has beenremoved for replacement. And if the laptop must be sent in for repair,the Latitude ON | FLASH module can be easily transferred to anotherLatitude ON Ready system while retaining the existing configurationand user information.Dell LatitudeON | FLASHoverviewThis video demonstrates thekey features of the LatitudeON | FLASH environment,including multiuserauthentication, the fullfeaturedWeb browser, andcomprehensive Virtual RemoteDesktop support.youtube.com/watch?v=zvda-cT69ekReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 97

Mobility solutions“Latitude ON | FLASHprovides a highly flexible,cost-effective way tomaximize productivityon the road.”Roberto Ayala is asolutions marketingmanager for the DellBusiness Client Group,where he developsFlexible Computingoutbound messaging.Steve Rokov isresponsible for enterprisemarketing at DeviceVM.DeviceVM partnerswith Dell to deliver thesoftware stack for LatitudeON | FLASH and LatitudeON | Reader.Secure business workspaceLatitude ON | FLASH was designed with security inmind. Because it provides a separate workspace,application changes to the Microsoft Windowsenvironment do not affect it, helping to reducesupport and security issues. IT staff can configuremultiple user accounts and passwords to helpensure that only authorized users gain access.Strong passwords are required for login.Latitude ON | FLASH uses a hardware firewalland is electrically isolated from the system harddrive and from the Windows OS, helping toprevent users and applications from accessingthe drive or the Windows partition. The lockeddownLinux ® OS–based environment is designedto ensure that applications cannot be installedor removed. Local user preferences arestored for each configured user, but browserplug-ins are not—helping reduce virus andmalware threats by preventing their payloadsfrom executing and installing themselves intothe permanent file system. The result is a clean,secure session each time the system boots intoLatitude ON | FLASH.Flexible remote desktop clientWith the Dell Virtual Remote Desktop (VRD)solution, the storage and processor functionality—the client “desktop”—is offloaded to a server andvirtualized, with the client device performing onlygraphics processing and user interface functions.Latitude ON | FLASH is well suited for this type ofclient virtualization, because it has pre-integratedCitrix and VMware remote desktop clients within alocked-down, secure OS along with an RDP clientfor Windows Terminal Services.The Citrix client supports Citrix XenDesktop and Citrix XenApp services using the ProgramNeighborhood and Web interfaces, includingsmooth roaming, client drive mapping, cut andpaste, and audio mapping functionality. TheVMware View client supports VMware virtualdesktop infrastructure software such as ThinApp,including USB redirection, Secure Sockets Layer(SSL), and direct mode functionality. The opensource rdesktop client is available to provideaccess to Windows Terminal Services using RDP.These options enable mobile and remote workersto turn their existing laptops into cost-effectivethin clients that can provide more flexibility thandedicated mobile thin-client hardware.Maximum efficiency and productivityThe Dell Latitude ON | FLASH module wasdesigned with the specific needs of mobile andremote workers in mind. By providing secure, nearinstantaccess to the Internet, Microsoft Officedocuments, virtual remote desktops, and more,it provides a highly flexible, cost-effective way tomaximize productivity on the road.Learn moreDell Latitude ON | FLASH:latitude-on-flash.comDell Latitude laptops:dell.com/latitudeDell virtual clients:dell.com/virtualclients98 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Windows ® . Life Without Walls .Dell recommends Windows 7.Where rugged works.Meet the rugged Dell Latitude E6400 XFR.T3 magazine called it one of the toughest gadgets on Earthand it’ll work in some of the toughest conditions on the planet.Whether that’s Alaska, Afghanistan, or anywhere in-between.If you need a computer that goes above and beyond the callof duty, visit Dell.com/ruggedE6400xfr or call 1-866-810-8610.Powered by Intel ® Core 2 Duo ProcessorIntel, the Intel Logo, Intel Inside, Intel Core, and Core Inside are trademarks of Intel Corporation in the U.S. and other countries.

Mobility solutionsSecuring the mobile workforcewith SonicWALL appliancesBy Patrick Sweeney and Matthew DieckmanThe mobile workforce presents significant security challenges to ITdepartments. SonicWALL ® Aventail ® E-Class Secure Remote Access(SRA) appliances are designed to support a comprehensive, policybasedapproach to remote access that helps streamline managementwhile enhancing security and flexibility.Mobile technology is changingthe face of business, enablingemployees to be productivefrom virtually anywhere, cuttingtime and costs out of operations, and helpingcompanies reach their customers effectively. Butthis technology is also presenting a significantchallenge to IT departments—especially whenit comes to security. With a growing range ofdevices and points of entry on the network, the“workplace” is no longer restricted to a protectedarea behind a hardened perimeter.In response, IT departments often findthemselves having to adopt a one-size-fitsallapproach to authenticating end usersand providing remote access, which often isnot enough in a multifaceted mobile world.Alternatively, they may create a thicket of individualrules to try to meet the different enterprise andsecurity needs presented by the wide range ofusers and devices, which can be complex, timeconsuming,and expensive.The growing emphasis on mobility requiresa sophisticated, nuanced view of security.SonicWALL Aventail E-Class Secure RemoteAccess (SRA) appliances are designed to support acomprehensive, policy-based approach to securityin Dell hardware–based IT environments—enabling access to mission-critical resources fromvirtually any endpoint device, streamlining policymanagement, and helping IT groups effectivelyprovide the security and flexibility to support theincreasingly mobile workforce.Blurring the network perimeterThe expansion of mobile technology means thatIT departments must contend with a wide rangeof data and voice devices, a growing number ofwhich are personal devices used by employees.Mark Bouchard, founder of the AimPointGroup research and analysis firm, calls this the“consumerization” of IT. IT groups, he writes, are“being given no option other than to support theongoing cross-over and melding of equipment,services, and activities of a personal nature withthose historically identified as being ‘businessoriented.’”As a result, “the IT department, alongwith its network security infrastructure, must beable to account for a mixed bag. Not all deviceswill have a robust security model, not all deviceswill be corporate-owned and managed, andneither will there be clear relationships between a100 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Employee workingon the roadRemote accessEmployee workingat a kioskEmployee usinga smartphoneDetectSonicWALL Aventail End Point Control technology continually detects theidentities and security states of endpoint devicesProtectSonicWALL Aventail Unified Policy technology enforces device access controlto help ensure that users access only authorized applicationsConnectSonicWALL Aventail Smart Access and Smart Tunneling technologies helpensure easy, secure user access to network resourcesEnterprise data centerEmployee workingfrom homeEmployee usinga wireless hotspotDetectProtectConnectDirectoriesExtranet accessInternetLANSonicWALL AventailE-Class SRA appliancesLightweight DirectoryAccess Protocol(LDAP) serverRSA AuthenticationManager serverApplicationsWebapplicationsClient/serverapplicationsFile sharesCustomer or supplierworking behind a firewallInternal accessInternal usersMicrosoftActive Directory ® serverDatabasesVoice-over-IP(VoIP)applicationsBusiness partnerusing a Web browserRemote AuthenticationDial-In User Service(RADIUS) serverFigure 1. SonicWALL Aventail E-Class SRA appliances help provide secure access across a comprehensive range of users, devices, and applicationsdevice’s location (e.g., inside thenetwork) and its type, securitystate, or category of user.” 1Beyond the use of mobiletechnology, other enterprisepractices are also blurring thetraditional network perimeter.Increased collaborationand the rise of extendedenterprises mean that a risingnumber of external partnersmay need to access internalapplications. And the use ofteleworking is growing—atrend that is likely to continue,as rising transportation costsand environmental concernsencourage flexibility regardingworking from home.Many organizations are alsolooking at the role that a remoteworkforce can play in businesscontinuity. This interest increasedsharply with the recent focus onthe H1N1 virus, which had thepotential to keep large numbersof people away from offices. Butdisruptions such as snowstorms,earthquakes, and terrorist actscould also require an increasedreliance on remote employees,and on the variety of managedand unmanaged devices theyneed to use.Perhaps most significantly,as the importance of remoteworkers grows, many companiesfind that the complexity ofproviding security in this typeof environment can actuallystifle growth and productivity.According to a report fromNemertes Research, “newapplications and servicescan, increasingly, transformoperations or customerinteractions, but 67% ofenterprises in NemertesSecurity and InformationProtection benchmark haverejected attractive and useful1 “Network Security for 2010 and Beyond – The Impact of the Consumerization, Webification, Virtualization, and Consolidation of IT,” by Mark Bouchard,AimPoint Group, February 2010, www.sonicwall.com/downloads/wp-eng-049_apg-network-security-2010-beyond.pdf.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 101

Mobility solutionstechnologies because of securityconcerns around them.” 2 Inessence, secure mobile accessis not just an IT concern, buta fundamental issue for theenterprise as a whole.Designing an effectivemobile security policySonicWALL Aventail E-ClassSRA appliances are designed tosupport an increasingly mobileworkforce securely and costeffectively.By using SecureSockets Layer (SSL) virtual privatenetwork (VPN) technologyrather than traditional IP Security(IPsec) VPN technology, theseappliances enable security atthe application level ratherthan the network level and canallow secure access even fromunmanaged devices. They alsoprovide a layered approach thatenables increased granularityof policy and access control,which is key when dealing witha variety of locations and remotedevices (see Figure 1).The SonicWALL AventailE-Class SRA policy-basedapproach encompasses threefundamental elements of mobilesecurity: user authentication,endpoint device identification,and application access. The firstelement, user authentication, isbased on authentication realms,which administrators set up toestablish how end users areidentified. Organizations cancreate multiple authenticationrealms to accommodate differentrequirements for different users.For example, some users maybe required to provide onlya username and password,whereas others may be requiredto use token authenticationor stacked authentication.E-Class SRA appliances supporta variety of authenticationmethods, including server-sidedigital certificates, usernamesand passwords, client-sidedigital certificates, RSA SecurIDand other onetime passwordtokens, and dual or stackedauthentication.For endpoint deviceidentification, E-Class SRAappliances let organizationsenforce granular access controlrules for Microsoft ® Windows ® ,Microsoft Windows Mobile ® ,Linux ® , and Apple Mac OS Xendpoints. Administrators cancreate device profiles thatinclude attributes such asthe presence of specific files,directories, registry settings, andantivirus programs; Windowshard drive serial number andWindows domain membership;and device type. These profilesare then used to scan and verifydevice attributes before accessis granted. Based on the resultsof the scan, the device is placedinto a security category, or“policy zone”; depending on thezone the device falls into, accessmay be granted or denied, orthe device may be quarantineduntil some action is taken,such as installing a requiredantivirus update. Policy zonesmake it possible to provide“SonicWALL Aventail E-Class SRAappliances are designed to supportan increasingly mobile workforcesecurely and cost-effectively.”differentiated access based onthe type of device being usedand the overall security stateof the device—a significantenhancement over a typicalapproach that grants accessto many potential resourcesregardless of the safety of theendpoint device.Finally, for applicationaccess, E-Class SRA applianceslook at the informationgathered about users anddevices—that is, the degreeof trust established for thosefactors—and then matches itto a customized authorizationpolicy that determines whichapplications and resources theuser and device can access. Forexample, the policy may specifythat employees working fromhome on unmanaged deviceswould be given just enoughapplication access to carry outthe basics of their work, whilethose same employees wouldhave greater access to moresensitive content when usinga highly secured, IT-manageddevice. A business partner usingan external corporate system,an IT technician working froma home PC, and a travelingexecutive with a companylaptop can all be given differentyet appropriate access withrelative ease—providing farmore flexibility than a typicalone-size-fits-all approach.The policy-based approachis not only effective; it alsohelps reduce the complexityof managing access. Thissimplicity is further enhancedwith the advanced, objectbasedSonicWALL AventailUnified Policy model, whichenables administrators toview and manage the entireorganization’s remote accesspolicy structure using a singleintegrated view in an intuitivemanagement console. Thecentralized administration isdesigned to consolidate controlof Web resources, file shares,and client/server applications ina single location.Unified Policy also enablesadministrators to quickly andeasily create a single rule setcovering a comprehensiverange of resources and accessmethods—avoiding the need tocreate separate access-control2 “The Center Is Everywhere,” by Andreas Antonopoulos and John Burke, Nemertes Research, February 2008, www.sonicwall.com/downloads/wp_nemertes_center-is-everywhere.pdf.102 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

ules for each user based onthe type of resource he orshe wants to access and theaccess method being used, allspread across separate pagesin the management interface.A single rule can define accessfor an entire group of usersacross multiple resources usingmultiple access methods. Forexample, from one consolepage, administrators can definea single rule granting everyuser in a sales group access tosales-specific Web resources,file shares, and client/serverapplications from theirworkstations or their managedsmartphones.Taking advantage of thelayered approachThe comprehensive, layeredapproach to mobile accessprovided by SonicWALL AventailE-Class SRA appliances canprovide a wide range of benefits,including the following:• Effective remote accesssecurity: Effective remoteaccess security includescomprehensive accessmanagement policies basednot only on users, but alsoon the level of risk presentedby the users’ environments.Granular access enablesorganizations to customizesecure access for diverseendpoints rather than relyingon a single approach that mayor may not be appropriate fora given device.• Streamlined IT management:Unified Policy helps to simplifyuse and administration, easethe burden on IT resources,and reduce systemsmanagement costs. Theobject-based model is alsohighly scalable—for example,E-Class SRA appliances havebeen successfully deployedin global organizations withmore than 85,000 users. Thisscalability helps to reducerepetition for administratorsand simplify network changes.• Low total cost of ownership:In addition to helping reducemanagement costs, E-ClassSRA appliances help lowerIT costs by enabling networkmanagers to easily deployand manage a single secureaccess gateway to networkresources for both internal andexternal users—including Webbrowser–based, host-based,and client/server applications.The appliances can beclientless or use lightweightWeb-delivered clients, whichhelps reduce managementoverhead and support calls.They can also deliver highavailability, with integrated loadbalancing and active/activefailover on some models,helping eliminate the need fora third-party load balancer.• Business continuity: Inthe event of a businessdisruption, organizationscan quickly establish secure,appropriate access to anexpanded remote workforce.E-Class SRA appliances canprovide a secure applicationaccess gateway at primarydata centers and serve as agateway to hot, warm, or colddisaster recovery facilities.An optional SonicWALLAventail Spike License Packlets organizations temporarilyand cost-effectively increasetheir remote user count tothe maximum capacity oftheir E-Class SRA appliance,whether this capacity is afew dozen or a few thousandadditional users.• Increased productivity:Employees and partnerscan have appropriateyet controlled access toenterprise applications froma wide variety of devices andlocations, helping to deliverthe right information to userswhen and where they need it.E-Class SRA appliances alsoprovide session persistence,which lets mobile users retaina current session when theyswitch between networks—forexample, as they movebetween offices, cars, homes,and hotels—without needingto re-authenticate.Supporting the mobileworkforceA mobile workforce brings withit specific security challengesthat call for comprehensive,policy-based control of remoteaccess. Using this type ofapproach in conjunction withSonicWALL Aventail E-Class SRAappliances and Dell hardware,IT departments can simplifythe management of remoteaccess without sacrificingsophistication or driving upcosts. And they can meet thedual demands of providinghigh levels of security whilesupporting the flexible deliveryof vital information—and helpthe enterprise thrive in the eraof the mobile workforce.Patrick Sweeney is vicepresident of the productmanagement team atSonicWALL, where heoversees its networksecurity, content security,business continuity, andpolicy and managementproduct lines.Matthew Dieckman isthe product managerfor SRA solutions atSonicWALL, where hedrives the developmentroad map for SonicWALLSSL VPN products andservices.Learn moreSonicWALL AventailE-Class SRA appliances:www.sonicwall.comReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 103

Business intelligenceEnabling pervasive businessintelligence using anintegrated Oracle platformBy Mark Conway, Dan Blankenship, and Jaime DelgadoAn integrated, standards-based platform offers an efficientway to move beyond the traditional, fragmented approach tobusiness intelligence. Taking advantage of Oracle ® BI softwareand Dell servers and storage can help weave cost-effective,analytics-driven decision making deep into the organization.Today, organizations are constantlycontending with increased volatility,complexity, and demands fortransparency and performance—and tokeep up, they must be armed with timely, accurateinformation. That reality has put the spotlight onbusiness intelligence (BI) and the analysis of data tosupport decision making.BI has been a top priority at many companiesfor years, offering a key way for senior executivesto gain fast, high-quality insights into performanceand operations. But the need for such analyticsdriveninsights has been growing at other levelsof the enterprise as well. As a result, there hasbeen a growing interest in “pervasive BI”—theuse of BI tools to share and analyze informationthroughout an organization. The idea is to gleaninsights from the vast amounts of data flowingthrough the organization, and then use them todrive predictions, decisions, and actions from theexecutive suite to the front line.Pervasive BI is clearly an appealing concept,but adoption has been relatively slow. As a recentsurvey report on “the intelligent enterprise” from theEconomist Intelligence Unit pointed out, “the abilityto make rapid, high-quality decisions based on themost timely, relevant and accurate informationavailable is critical and a common characteristicof the world’s leading companies. Yet as thissurvey shows, few have developed the necessaryprocesses for sharing and analysing criticaloperational information across the enterprise.” 1 Toa large extent, this situation is due to the traditionalfragmented approach to BI. Organizationsoften have multiple BI and analytical tools andarchitectures deployed in various departments.This fragmentation leads to siloed views, makingit difficult to share data and insights across theorganization, and often creating inconsistencies inreporting. It also leads to inefficiencies in reconcilingvarious data points and reports, or, worse yet, keydecisions being made based on inconsistent data.The solution is a comprehensive approach thatembeds standardized enterprise BI platforms andanalytics across the organization. Such an approachoffers the advantages that typically come with1 “The Intelligent Enterprise: Creating a Culture of Speedy and Efficient Decision-Making,” by the Economist Intelligence Unit, December 2009, available ateiu.com/site_info.asp?info_name=oracle_decision_making.104 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

InteractivedashboardsReportingand publishingAd hoc queriesand analysesDetectionand alertingMobile andoffline accessMicrosoft OfficeintegrationSearchEmbeddeddataStrategicplanningFinancialclosePlanning andbudgetingProfitabilitymanagementPervasive information accessERPanalyticsCRManalyticsIndustryanalyticsOracle BI ServerCommon analytic foundationOracle EssbaseFigure 1. IntegratedOracle platform forpervasive businessintelligenceOnline transactionprocessing (OLTP)and operationaldata store (ODS)systemsData warehouseand data martsOLAP systemsOracle, PeopleSoft,JD Edwards, Siebel,SAP, and customsystemsMicrosoft OfficeExcel and XMLBusinessprocessesstandardization, including increased efficiencyand consistency as well as reduced complexity.That perspective is at the heart of Oracle’sapproach to BI. The Oracle enterprise BIplatform encompasses a comprehensiverange of BI tools and technologies. But,more important, it integrates these toolsand technologies to create a platform that ismore than the sum of its parts. Organizationscan take advantage of this BI platform aswell as Dell servers and storage to create ascalable, enterprise-class approach to BI andultimately drive fact-based decision makingand increased business agility.Oracle business intelligenceplatform componentsTo help meet the BI needs of organizationsof all sizes, Oracle has made significantinvestments in this area—including theacquisition of Hyperion and its financialanalyticsand performance-management tools,and the acquisition of Siebel and its extensivesuite of analytical technologies. Oracle hasfocused on the integration and standardizationof BI tools using open technologies to createa true enterprise BI platform that supports allthree of the fundamental types of BI—givingorganizations the query and reporting toolsneeded to understand their business, theanalytic tools needed to pose a range ofquestions, and the modeling tools neededto look ahead, manage risk and uncertainty,and exploit opportunities. With this platform,Oracle was recently recognized by IDC as thelargest vendor in business analytics in 2008. 2 InJanuary, Oracle was positioned as a “leader” inGartner’s BI Platform Magic Quadrant. 3The Oracle platform is designed tosupport pervasive BI with a unified foundationthat includes three fundamental components(see Figure 1):• Oracle BI Server: Oracle BI Server isa scalable, sophisticated data access,aggregation, and calculation enginefor creating real-time intelligence fromheterogeneous data sources—includingOracle, Microsoft ® SQL Server ® , IBM ®DB2, and Teradata software—anddelivering it efficiently across even largescaleenterprise environments. It cansemantically integrate information and,through a combination of a distributedquery engine and a logical, semanticrepresentation of the data, can enableorganizations to represent multiplephysical data sources as a single,simplified data structure to end-user tools.This simplified view can then be accessed2 “Worldwide Business Analytics Software 2009–2013 Forecast and 2008 Vendor Shares,” by IDC, Doc #219383, August 2009.3 “Magic Quadrant for Business Intelligence Platforms,” by Rita L. Sallam, Bill Hostmann, James Richardson, and Andreas Bitterer, Gartner, Inc., January 29, 2010. The Magic Quadrant is copyrighted 2010 byGartner, Inc. and is reused with permission. The Magic Quadrant is a graphical representation of a marketplace at and for a specific time period. It depicts Gartner’s analysis of how certain vendors measureagainst criteria for that marketplace, as defined by Gartner. Gartner does not endorse any vendor, product, or service depicted in the Magic Quadrant, and does not advise technology users to select onlythose vendors placed in the “Leaders” quadrant. The Magic Quadrant is intended solely as a research tool, and is not meant to be a specific guide to action. Gartner disclaims all warranties, express or implied,with respect to this research, including any warranties of merchantability or fitness for a particular purpose.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 105

Business intelligence“For IT departments, anenterprise BI platformhelps to streamline BImodeling, reduce the needfor support, and simplifychanges and upgrades.”Deployingpervasive BIImplementing true pervasiveBI can present significantdifficulties for organizationscontending with complexbusiness processes, data silos,and other challenges. In thispodcast, BI experts from Delland Oracle share insightson trends and best practicesthat organizations can begintaking advantage of today.oracle.com/goto/bipodcastby the end-user tools through a standard SQLbasedinterface.• Oracle Essbase: This online analytical processing(OLAP) server software provides an environmentfor the rapid development of custom analyticand enterprise performance-managementapplications. Based on a high-performancecalculation engine, it provides rapid responsetimes and an intuitive interface, enabling users tointeract with the data quickly without requiringsupport from technical experts—helping them toanalyze the metrics that influence performanceand enhance their decision making. (For moreinformation, see the “Thinking ahead with OracleEssbase” sidebar in this article.)• Oracle Real Time Decisions (RTD): OracleRTD is a high-performance transactional serverplatform and a predictive analytics engine thatlets organizations “embed” analytics and BIinto operational processes. Using a serviceorientedJava 2 Platform, Enterprise Edition(J2EE) architecture, it is designed to providereal-time decisions and recommendations tohelp organizations align frontline actions withperformance goals. Oracle RTD can supportdecision logic based on user-defined rules orprovide automatically managed self-learningpredictive models, enabling organizations tolearn from each interaction and adjust theirprocesses in real time.A key element in the Oracle BI approach is itscommon enterprise information model, a unifiedmetadata model that consists of three tiers: aphysical layer that allows organizations to importthe table structures of their existing data sources,a semantic object layer that lets them create asimplified semantic representation of multiple datasources and a model of the business that reflectsthe end user’s perspective, and a presentation layerthat exposes this business model to end users.This common enterprise information modelis accessed by Oracle analytics applications andend-user BI tools, giving each user and departmentthe same consistent view of information,customized to their needs. Unlike the traditionalapproach, which typically requires organizationsto use different metadata for their ad hoc tools,reporting tools, and dashboard tools, the Oracleapproach avoids the need to maintain multiplemetadata environments for different types of users.Organizations can instead “model once, deployeverywhere”—that is, they can define key metricsand calculations in one place, and then use thatbase to deliver a consistent view of informationacross the organization to support efficient,effective decision making.Comprehensive toolsRunning the Oracle BI foundation on Dellservers and storage provides a unified enterpriseinfrastructure for a comprehensive suite of BI andanalytics tools. These tools provide organizationswith a variety of key capabilities, includingthe following:• Production reporting: The Oracle BI Publisherintegrated reporting component enablesthe creation of highly formatted templates,reports, and documents, such as checks andgovernment forms, in complex and distributedenvironments. The Oracle architectureenables organizations to generate and deliverinformation to employees, customers, andpartners securely and in the appropriate format.End users can work with familiar tools such asMicrosoft Office Word or Adobe ® Acrobat forreport layouts.• Interactive dashboards: A thin client solutionprovides interactive collections of analytic106 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Thinking ahead with Oracle EssbaseDecision makers are constantly trying to determine what is likelyto happen in the future—and pervasive business intelligence (BI)can help users across the organization do just that.Oracle Essbase puts powerful analytical tools directly intothe hands of end users. Among other features, it enables theseusers to perform forward-looking analyses that model complexscenarios and create forecasts. They can use these tools toidentify patterns, such as determining which customers spendthe most, and to explore questions such as “What if demandfalls short of forecast?” or “What if we rolled out this productnationwide?” The ability to interact quickly and easily with thedata—as opposed to reading through static reports—helpsincrease user adoption of the system and supports rapid,focused decision making.At its core, Oracle Essbase contains a high-performancecalculation engine with more than 350 prebuilt functions.Oracle Essbase also provides a rapid application developmentenvironment that is managed by the business, enabling users toquickly create applications for querying and reporting, analytics,and forward-looking modeling. It also enables pervasive BI bysupporting extremely fast query response times for vast numbersof simultaneous users, large data sets, and complex models.Oracle Essbase also helps simplify enterprise BI support forIT departments. It can be configured, deployed, and maintainedby the business organization, and its n-tier architecture providesconnection pooling, load balancing, and automatic failoverto help meet service-level requirements. Because it workswith both 32- and 64-bit computing, it is capable of massivescalability for enterprise-wide deployments. Its support for64-bit architectures also enables larger analytical models withshorter calculation times than would be possible in 32-bitenvironments, helping increase the potential size of analyticapplications and the number of concurrent users.Meanwhile, user scalability features such as caching,multi-threading, partitioning, and cross-platform supportenable IT professionals to use a reduced number of serversto support many analytic applications and large usercommunities. Oracle Essbase also includes a number ofmanageability features, such as Microsoft .NET developmenttools, certified enterprise resource planning (ERP) andcustomer relationship management (CRM) applicationintegration adapters, administrative wizards, and automatedmaintenance scripts—all of which help reduce the deploymentand maintenance costs of pervasive BI.content with a rich variety of visualization tools.Running on a Web-based architecture, thesedashboards provide end users with informationfiltered and personalized for their specificidentity, function, or role based on predefinedsecurity rules. The interactive interface helpsmake the presentation of data intuitive, relevant,and easy to understand, while guided navigationand alerts help users gain insight and takeaction. Making information personalized andproviding it in the appropriate context is oneof the keys to encouraging user adoption andmoving organizations along the pervasive BImaturity curve.• Ad hoc queries and analyses: End users cancreate new analyses or modify existing analysesin dashboard pages. To help free users fromdata structure complexity, the metadata layerof Oracle BI offers a logical view of metrics,hierarchies, and calculations expressed inunderstandable concepts. End users do notneed to understand physical data storageto combine data from multiple enterpriseinformation sources.To further drive pervasive BI, Oracle BIApplications provide prebuilt applications thatdeliver intuitive, role-based intelligence throughoutthe organization—from frontline employees tosenior management. Broadly speaking, theseapplications fall into two categories: customerrelationship management (CRM) applicationssuch as sales or customer analytics software, andenterprise resource planning (ERP) applicationssuch as financial, supply chain, or humanresources analytics software. They each includefour key components: a prebuilt warehouse; premappedmetadata embedding best practices formetrics and key performance indicators; prebuiltextract, transform, and load (ETL) capabilities toReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 107

Business intelligenceAccelerating business intelligence with Dell ServicesDell Services uses a holistic, collaborative approach to businessintelligence (BI) analytics that incorporates information frommultiple sources to give management comprehensive reportingon financial and operational performance. The Dell Servicesapproach is integrated across the enterprise to deliver broadbasedinsight into overall enterprise efficiency, with deep-diveanalytics that enhance perspective and decision making.The Dell Optimized Business Intelligence Analyticsprogram integrates key financial and operationalinformational analytics throughout the organization—with afocus on using best practices to standardize processes anddrive profitability. The program is an accelerated enterprise BIprocess that blends technology and business processes usingthe Oracle BI framework as an accelerator to help consolidateand deliver financial, human capital, supply chain, and otheroperational performance measures. The program is designedto take full advantage of Oracle BI Applications to extract andconsolidate data from enterprise resource planning (ERP)systems in a single process framework, enabling enterprisewideorganizational intelligence.Using Dell servers and storage along with Oracle BItools—and drawing on prebuilt configurations and solutions,including installing the Informatica PowerCenter data load tool,Oracle BI Server, and Oracle analytics applications, offeredthrough the partnership of Dell and Oracle—Dell Services canhelp organizations quickly build solutions and configurationscustomized to meet their needs and enable them to transforminformation into actionable knowledge.extract data from Oracle and non-Oracle sources;and a best-practice library of dashboards andreports. They also include guided analytics thatwalk users through analyses using straightforward,natural questions. These BI applications helporganizations shorten time to value whileextending BI throughout the organization, using asingle common foundation. Organizations can alsotake advantage of Dell Services to help optimizedeployments of the Oracle platform and OracleBI Applications in their own environments (seethe “Accelerating business intelligence with DellServices” sidebar in this article).Simplified integrationTo be effective, a BI platform naturally needs tointegrate easily with other systems. To that end,the Oracle BI platform helps reduce the integrationburden on IT groups in several ways. For example,Oracle BI tools are pre-integrated with one anotherand with Oracle transactional applications such asOracle E-Business Suite, PeopleSoft ® Enterprise,JD Edwards ® EnterpriseOne, Siebel CRM, andOracle CRM On Demand software.As an open platform, Oracle BI also uses a“hot-pluggable” strategy to help integrate withnon-Oracle technologies and applications. TheOracle platform is designed to easily integratewith existing IT infrastructures and data sources,including ETL tools, application servers, securityinfrastructures, and other front-end and analyticaltools. The Oracle BI platform can access datafrom multiple heterogeneous sources—includingrelational databases such as Oracle Database,Microsoft SQL Server, IBM DB2, and Teradatasoftware; multidimensional sources such asOracle Essbase, SAP ® NetWeaver ® BI, andMicrosoft Analysis Services software; flat files;XML data; and unstructured data. It also integrateswith standard portals, Web services, andauthentication technologies.End users are, of course, a key part ofpervasive BI, so this focus on integration alsoextends to the desktop. An Oracle BI plug-in canintegrate BI information from Oracle BI Server andOracle reporting tools into the Microsoft Officeenvironment, enabling users to embed up-to-theminutedata in Microsoft Office Word, Excel ® , andPowerPoint ® documents.Business intelligence acrossthe organizationThe Oracle approach adds up to having oneplatform and one information model supportingmultiple delivery channels, helping to enhancedecision making and increase agility throughout108 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

“Unlike the traditional approach,which typically requires organizationsto use different metadata for theirad hoc tools, reporting tools, anddashboard tools, the Oracle approachavoids the need to maintain multiplemetadata environments for differenttypes of users.”the organization. Today, theneed for that kind of capabilityis not restricted to largeenterprises—organizations ofall sizes can benefit from anenterprise approach to BI.To support the needs of bothlarge and small organizations,the Oracle platform is availablein two versions. Oracle BI SuiteEnterprise Edition is designed forlarge organizations; in additionto core capabilities such asinteractive dashboards and adhoc queries and analyses, thissuite incorporates features suchas proactive intelligence andalerts as well as disconnectedanalytics, which allow mobileusers to use BI tools offline. Italso bundles a number of keyOracle Hyperion financial andWeb analysis tools.Oracle BusinessIntelligence Standard EditionOne, meanwhile, is basedon Oracle BI Suite EnterpriseEdition technology but ispackaged and preconfiguredfor deployments of 5–50 users.Designed to run on one server,it provides a comprehensiveand integrated BI system, andincludes Oracle BI Server aswell as interactive dashboardsand support for highlyformatted reporting and ad hocreports and analyses. It alsoincludes Oracle Database andpowerful ETL tools for buildingdata warehouses.Both versions offer similaradvantages. For IT departments,an enterprise BI platform helpsto streamline BI modeling,reduce the need for support,and simplify changes andupgrades. For the organizationas a whole, this platformprovides a simplified way toprovide consistent, accurateinformation to drive decisions;deploy a variety of front-end BItools without creating BI silos;and move toward pervasiveBI. Because both versions arebuilt on the same technology,organizations can easilyexpand their BI capabilities asneeded. By taking advantageof this seamless scalabilityalong with the wide rangeof available Dell servers andstorage, organizations canstart small and then scale theirdeployments later as theirrequirements increase.Many organizations stillhave a ways to go in achievingpervasive BI, and the fieldcontinues to evolve. Ongoingenhancements are likely tofurther improve the ability tolink insight to business actions,and to give end users accessto advanced visualization andother tools to support effectivedecision making. Having anopen, integrated, standardBI platform that can be usedacross the organization is keyto pervasive BI, and continuesto be key to the continuedevolution of BI.Mark Conway is thedirector of productmarketing for businessintelligence at Oracle,where he orchestratesBI marketing strategyand global marketinginitiatives.Dan Blankenship is thebusiness intelligencepractice leader for DellServices.Jaime Delgado isa senior director ofworldwide alliancesat Oracle, where he isresponsible for the DellGlobal Alliance.Learn moreOracle Business Intelligence:oracle.com/appserver/business-intelligenceOracle and Dell:dell.com/oracleoracle.com/dellReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 109

OEM solutionsEnabling eco-friendlydeployments with NEI platformsBy Richard GraberFor organizations planning to build a new data center orexpand a suitable existing facility, implementing a DC-basedpower infrastructure can offer significant advantages. Basedon the Dell PowerEdge R710 server, NEI’s E-2710 providesa cost-effective, purpose-built application platform designedto maximize efficiency in these environments.Gauging data centerefficiencyPUE =1DCiE = =PUETotal facility powerIT equipment powerIT equipment powerTotal facility powerPower consumption has become an increasinglyimportant topic for enterprise data centers, particularlyas the economy has forced organizations to focuson ways to maximize energy, performance, and costefficiencies. Rising power and cooling costs, meanwhile, haveresulted in IT departments investigating ways to take advantageof eco-friendly technologies that can help increase resourceutilization, availability, and compute density while also helpinglower the costs associated with power, cooling, and maintenance.Servers are one of the biggest power users in data centers today,which explains why virtualizing x86-based servers is among themost popular energy-saving strategies.Globally, governments are also putting their weight behindeco-friendly technologies. Data centers typically require vastlymore energy per square foot than a typical office building, andwhile data center consolidation has remained a primary focus,some administrative bodies have also turned their attention toenergy conservation and power management in operating systemsand computing hardware. In May 2009, for example, the U.S.Environmental Protection Agency (EPA) put new ENERGY STARrequirements for enterprise-class servers into effect; other U.S.agencies like the Department of Energy (DOE) are also workingwith industry associations to prepare standards and create tools tohelp measure and enhance data center efficiency. This efficiency iscommonly defined by the Power Usage Effectiveness (PUE) and DataCenter Infrastructure Efficiency (DCiE) metrics shown on the left.110 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

NEI’s E-2710The E-2710 provides a highly efficient,power-optimized, standards-based platformfor eco-friendly deploymentsHot-swappable hard drives andredundant power supplieshelp maximize E-2710 reliabilityand uptimeA data center having a PUE value of 2.0, forexample, would draw a total of 2 W of power forevery 1 W consumed by IT equipment. This samedata center would have a DCiE value of 0.5—aninverse way of presenting the same measure. Aperfectly efficient data center, in which 100 percentof the power was used directly by IT equipment,would have PUE and DCiE values of exactly 1.0.Among the many technologies that can helpimprove data center PUE and DCiE is a smart,end-to-end power architecture. In some datacenters, one option that can help maximizeefficiency is an infrastructure that converts ACpower from the utility to DC power inside thedata center itself. In the short term, this type ofdesign is not likely to replace existing structures,particularly in recently constructed facilities thathave not yet reached payback on their initialinvestments and in facilities being retrofittedor upgraded. For organizations planning tobuild a new facility or to significantly expand asuitable existing facility, however, moving toa DC-powered infrastructure can potentially offersignificant advantages. This article examinesthe requirements of an end-to-end DC-baseddata center and introduces an advanced, highlyefficient DC-powered platform from Dell andNEI that is designed to fit seamlessly into theseinfrastructures—the E-2710.Evolution of DC-powered data centersThe evolution toward DC-powered data centershas been proceeding for some time. DC isalready the resident power used by data centersubsystems, primarily at the microchip andboard level, and virtually all active and passivesemiconductor integrated circuits operate fromlow-level DC voltages. High-speed processors,memory and storage devices, and I/O circuitry allconsume DC power, which is readily regulatedand distributed in x86-based servers.Lower and lower DC voltages are alsobeing used to power microcircuits, and dataprocessing performance has become increasinglyefficient with every successive generation ofprocessors. The Intel ® Xeon ® processor 5500series in 11th-generation Dell PowerEdge servers,for example, incorporates power-throttlingtechnologies designed to significantly increasememory bandwidth and performance per wattover previous-generation processors. The trendhas been to expand native DC efficiencies beyondthe subsystem level and eliminate multiple (andoften redundant) power conversion steps tohelp increase overall efficiency within the ITinfrastructure. In terms of adopting DC powerin the data center, the servers and data centerappliances themselves are generally no longerthe conversion break point.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 111

OEM solutionsAC-based infrastructure750 kVA UPSs 150 kW PDUsAC-to-DCpower suppliesServersFigure 1. Equipping adata center with highvoltageDC power caneliminate inefficient powerconversion stepsHigh-voltageAC power2.5 MVA MV/LVtransformer480 V AC 480 V AC 208 V AC 48 V DCUtilityHigh-voltageAC powerDC-based infrastructure120 kWconverters–575 or Mainframe–400 V DC2.5 MVA MV/LVrectifier48 V DCServersEfficient design and the powerconversion processA typical data center includes numerousconversions from high-voltage AC powerto multilevel DC power. The powerconverters that perform these conversionsand regulate downstream componentsare more efficient than they used tobe, but are still typically expensive topurchase, service, and maintain over theirlife spans. Supplying the data center withhigh-voltage DC power rather than ACpower can help minimize or eliminatethese conversion steps (see Figure 1).Reducing the number of componentscan also translate to increased efficiencyand reduced heat dissipation—helpingreduce costs and water usage associatedwith server room cooling—as well as asimplified overall design.The energy efficiency of DC systemsis a measurement of the end-to-endsystem efficiency, not merely a singlecomponent within the system, which ishow organizations pay for power. Using–575, –380, and –48 V DC power distributioncan provide substantial advantages insome data centers—including increasedenergy efficiency, reliability, scalability,and safety; reduced space requirements,installation costs, and maintenance costs;simplified integration of renewableenergy; and eligibility for utility rebatesand credits in some areas. In the UnitedStates, Energy Efficiency Certificates (EECs)and Renewable Energy Certificates (RECs)may also be available, helping make DCpower a viable long-term choice for theorganization as a whole.Because many renewable energysources generate DC power, using theseenergy sources helps avoid the need toadd multiple conversions to accommodatean AC power path—helping to eliminateinverters and save significant amounts ofpower. And because DC-based data centerstypically require fewer power componentsthan comparable AC-based data centers,such facilities can potentially be more reliable,simply because there are fewer pieces to fail.DC systems also do not need to synchronizesine waves or frequencies across multiplesources, which can help to furthereliminate points of failure and simplifythe system.With fewer components to installand maintain and less heat to dissipate, aDC-based data center’s density per squarefoot can also potentially be greater than atraditional AC-based data center design.Bulk rectifiers can be sized up to 2.5 MWin a single unit, enabling denser powerfactors compared with AC topologies,and the resulting savings on space alonecan be substantial over the life of a typicalsystem. DC distribution systems arealso more readily designed to accommodatemodularity compared with typicalAC systems, helping support a scalableinfrastructure and growth strategy.In addition, in conjunction with simplepower electronics and appropriate DCcircuit breaker technology, DC systemscan be safer than AC systems from 380 Vto 600 V. When using 48 V DC power, forexample, the system is considered saferfrom a handling perspective than 110 V ACor 220 V AC systems.112 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Server platform for theDC-powered data centerOne of the difficulties withmoving from high-voltageAC power to DC power is thegeneral lack of IT equipmentdesigned to handle DC power.To help address this gap, Delland NEI have introduced theE-2710—a highly efficient,power-optimized platformbased on the Dell PowerEdgeR710 server. Available as anAC- or DC-powered platform, theE-2710 is purpose-built to helpmeet the needs of enterprise andtelecommunications applicationdevelopers and originalequipment manufacturers(OEMs) worldwide, as part ofeco-friendly deployments indata centers and carrier-classcentral offices that do notrequire Network Equipment-Building System (NEBS)Level 3 certification.This highly available,standards-based platformfeatures up to two quad-coreprocessors from the Intel Xeonprocessor 5500 series to helpdynamically scale power usageto specific workloads and helpincrease energy efficiency. Itcan also host NEI Smart Servicesto deliver network-ready, lifecycle-managedapplications;OS and patch management;streamlined support processes;and accelerated deployments.NEI offers support anddeployment services aswell as reverse logistics forthis platform.The E-2710 is also designedto be highly flexible andcustomizable while providingcost-effective, scalable, andreliable performance. In a 2Urack-mount design, it supportsup to four high-end telephonyPCI Express (PCIe) cards andup to eight Serial AttachedSCSI (SAS) or Serial ATA (SATA)hard drives—enabling OEMs tocustomize the system for specificstorage and I/O requirements.Integrated hardware RAID,hot-swappable hard drives, andredundant hot-swappable powersupplies help maximize reliabilityand uptime. The E-2710 is alsocompatible with the IntegratedDell Remote Access Controller(iDRAC) for remote powermanagement, virtual mediaaccess, and other console-drivencapabilities, all available througha supported Web browser.Optimized, eco-friendlydata center designFor some data centers,implementing a DC-basedpower infrastructure can bringsubstantial benefits—rangingfrom reduced cooling coststo simplified maintenance toincreased safety and reliability.Designed to fit seamlessly withinthese types of environments,the E-2710 provides a flexible,purpose-built platform forenterprise data centers andOEMs that can help maximizeefficiency while supportingcomprehensive eco-friendlystrategies.Learn moreNEI and Dell:www.nei.com/dellRichard Graber is seniorvice president of engineeringand operations at NEI.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.Learn how new DC-powered Dell TM PowerEdge TM servers canaddress the need for more energy-efficient ways to runenterprise-class applications and data center operations.Free whitepaper: Enabling Greenfield Application Deploymentswith PowerEdge TechnologiesFor more information or to download the whitepaper:visit www.nei.com/dell or call +1.972.633.3400

OEM solutionsCustom fit:The Dell PowerEdge R310appliance server platformBy Franklin FlintThe Dell PowerEdge R310 server offers enterprise-classfeatures in a high-performance, one-socket, 1U rackserver that is well suited for original equipmentmanufacturers (OEMs). Its small form factor and widerange of options provide a versatile, OEM-readyplatform for custom-built appliances.Dell PowerEdge R310 appliance serverBy offering flexibility, uptime,connectivity, and robust managementfeatures, the Dell PowerEdge R310can help meet a breadth ofOEM appliance requirements.When original equipment manufacturers (OEMs)develop custom appliances, they often havespecific requirements for the server they use asa platform. For example, appliances for certainmedical applications might require extremely high availability toavoid downtime or the need for IT intervention. Other appliances,such as those for networking and digital media applications, mayrequire more connectivity options and bandwidth than manystandard server applications. Still others—such as those that residein shallow and legacy racks, mobile server enclosures, and othernonstandard locations—may require a compact form factor.The 11th-generation Dell PowerEdge R310 is a single-socket,1U rack server designed specifically with OEMs in mind. Availableas either a standard Dell-branded server or an unbranded,OEM-ready appliance platform that can be re-branded by OEMs, thishighly flexible system combines cost-effective, high-performancecomputing power and a small form factor with a range ofconfigurable uptime, connectivity, and management featuresdesigned to meet a breadth of custom appliance requirements.Additionally, Dell offers a range of services designed to supplyOEMs with cost-effective, reliable appliance server platforms thatcan easily be custom configured and custom branded.114 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

Combiningcost-effectiveness withperformance and reliabilityOEMs have a range of specificrequirements, so to helpmeet these needs, the DellPowerEdge R310 platform offersa broad range of performanceand functionality options,and can scale from basic,nonredundant configurationswith limited manageability tohighly redundant configurationswith enterprise-class systemsmanagement capabilities (see the“Ease of management” sidebarin this article). Processor optionsrange from the entry-level Intel ®Celeron ® processor to thehigh-end Intel Xeon ® X3400series processor, while the sixdual in-line memory module(DIMM) sockets can provide upto 32 GB of memory capacity.Four hard drive bays are availablefor either 2.5- or 3.5-inch cabledor hot-swappable hard drives,and include support for SerialAttached SCSI (SAS) drives, SerialATA (SATA) II drives, and solidstatedrives (SSDs), for a total rawcapacity of up to 8 TB.The PowerEdge R310also offers full-height andhalf-length PCI Express (PCIe)expansion slots with x8 andx16 connectivity, a dedicatedstorage controller PCIe slot, andan embedded dual-port GigabitEthernet network interface card(NIC). The system supports theMicrosoft ® Windows ® , Red Hat ®Enterprise Linux ® , Novell ® SUSE ®Linux Enterprise Server, andSun Solaris operating systems;custom services can providecompatibility with less commonoperating systems.Ease of managementThe Dell PowerEdge R310 incorporates multiple features designed tosimplify and streamline management.Embedded Dell Lifecycle Controller: This optional feature performsprovisioning functions such as systems deployment, updates, hardwareconfiguration, and diagnostics from a single intuitive interface. TheLifecycle Controller is OS independent, helping avoid the need forpatches and updates to be compatible with the OS.Dell Management Console powered by Altiris from Symantec :This unified console software helps simplify systems managementby providing IT administrators with a consolidated view of their ITinfrastructure. Dell Management Console is included at no additionalcost with all Dell servers.Interactive LCD: This optional feature enables state-of-the-artserviceability and diagnostic capabilities.Integrated Dell Remote Access Controller 6 (iDRAC 6) Enterprise:This optional out-of-band management controller enables remotecontrol, event diagnosis, updates, and other management features.High uptime andextensive connectivityMany of the features availablein the PowerEdge R310 aredesigned specifically for uptimeand connectivity—factorsthat are often paramount forOEM custom appliances. Forexample, appliances that havehigh-availability requirementscan leverage the fact thatthe PowerEdge R310 can beconfigured with options thatinclude redundant powersupplies; hot-swappable harddrives, including hot-pluggableSAS and SATA II drives andSSDs; a range of RAID storagecontrollers from low-end,software-based chipset RAID upto high-performance RAID withadvanced battery-backed cache;and front-mounted status LCDpanels. Comprehensive remotemanagement capabilities andremote access capabilitiesare also available through theIntegrated Dell Remote AccessController (iDRAC).For appliances with highconnectivity requirements, theconnectivity features includedwith the PowerEdge R310 offereasily expandable I/O bandwidththrough two PCIe expansion slots,a dedicated storage controllerPCIe slot, and two GigabitEthernet NICs. This combinationof expansion slots and embeddedNICs is designed for dramaticallyenhanced connectivity withina 1U rack server, making thePowerEdge R310 well suited forI/O-heavy appliance softwaresuch as video, networking, andvoice-over-IP (VoIP) applications.Reprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.dell.com/powersolutions | 2010 Issue 02 115

OEM solutions“Many of the features available in thePowerEdge R310 are designed specifically foruptime and connectivity—factors that are oftenparamount for OEM custom appliances.”Generic bezels and faceplatesenable fast, simple custombranding for OEM-readyDell PowerEdge R310 appliancesIn addition, to support custom appliancesthat may be deployed in space-limitedareas, the PowerEdge R310 has a chassisdepth of only 24 inches—enabling it to fit innonstandard locations such as shallow andlegacy racks and mobile server enclosures.FlexibilityThe PowerEdge R310 is designed forextensive flexibility, enabling OEMs to offerplatforms with custom configurationsto help precisely meet performance,uptime, connectivity, and managementrequirements. OEMs can select from severalconfigurations for different use cases—forexample, production appliances might beconfigured for maximum performance andavailability, while test and developmentappliances may use cost-effectiveconfigurations. With this flexibility, OEMspay for only the features they need whilemaintaining room for growth.Delivering an OEM-ready platformTo facilitate development of customappliances, the Dell OEM Solutions teamoffers a range of services designed todeliver OEM-ready appliances. Theseappliances are based on standard Dellhardware platforms—including theDell PowerEdge R310 appliance serverplatform—that have been de-brandedto allow fast, simple customization.OEM-ready platforms incorporate a rangeof features, including the following:• Generic bezels and faceplates: Genericbezels allow for custom OEM branding.Bezel design templates such as the BezelDesign Kit are available at no additionalcost and allow OEMs to easily engineertheir own bezels for a custom look andfeel. Dell OEM Solutions offers custombezel design services as well.• Unbranded BIOS splash screen: OEMreadyservers ship with a generic splashscreen.• Generic packaging: OEM-ready serversare delivered with unbranded packaging,including no logos; unbranded regulatory,safety, and environmental literature; andplain packing tape.• Unbranded Unified Server Configurator(USC): The USC performs provisioningfunctions through the Dell LifecycleController, and is available unbrandedon OEM-ready platforms. The platformupdate feature built into the LifecycleController has been removed to helpprevent Dell-sanctioned product updatesthat may not be approved by the OEM,and that could otherwise potentiallycause issues with the applicationsrunning on the server platform.Dell also offers a range of optionalOEM-ready services to further facilitate therapid, easy development of OEM appliances,including custom integrations, customshipping, and custom inventory services.Powering custom appliancesThe Dell PowerEdge R310 applianceserver platform can deliver outstandingperformance, reliability, and connectivityin a compact, cost-efficient, single-socketrack server. This versatile system is wellsuited for custom OEM appliances designedfor a wide range of applications, includingdigital media, networking, and applicationsin medical and health care, retail, and otherspecific environments. OEM-ready servicesfrom Dell OEM Solutions allow standardPowerEdge R310 servers to be quickly andeasily customized to help OEMs streamlinedevelopment of applications on customappliance server platform configurations.Learn moreFranklin Flint is a globalenterprise technology evangelistfor OEM Solutions at Dell.Dell OEM Solutions:dell.com/oemDell PowerEdge appliances:dell.com/appliances116 2010 Issue 02 | dell.com/powersolutionsReprinted from Dell Power Solutions, 2010 Issue 2. Copyright © 2010 Dell Inc. All rights reserved.

In large enterprises, IT running costs devour up to80% of IT budgets, starving innovation of funds.Break the status quo with Dell. Our open, standardsbasedsolutions let you scale your infrastructure easily,and within the realities of your budget. So you caninvest more in innovation and let the status quo starvefor a change. Would you like to make your companya more Efficient Enterprise?Reinvest in innovation @ dell.com/efficiententerprise

MidsizeIndustryLeadersOver 5,000 Midsize Consumer Goods CompaniesOver 20,000 Midsize High Tech CompaniesOver 36,000 Midsize ManufacturersOver 5,000 Midsize BanksGet Better Results With Oracleoracle.com/goto/midsizeor call 1.800.ORACLE.1Copyright © 2009, Oracle. All rights reserved. Oracle is a registered trademark of Oracle Corporation and/or its affiliates.Other names may be trademarks of their respective owners.