CHAPTER 1

VIRTUALIZATIONCONCEPTSINDISASTERRECOVERY

Virtualization is changingthe disaster recovery

landscape. Rethinking yourhardware and software, testingmethods and performancerequirements is key.

p2EDITOR’S NOTE

p3VIRTUALIZATIONCONCEPTS IN DR

p4PROS AND CONS

p6INFRASTRUCTURE

p8CASE STUDY

p 10TESTING ANDPERFORMANCE

Introduction to Virtualization E-book

UPHEAVAL AND CHANGE are par for thecourse in the IT sector, but occasion-ally, a new technology comes alongthat profoundly changes the game.Virtualization is one such technology.Virtualization lets IT departments

decouple their services and resourcesfrom the underlying hardware forgreater infrastructure flexibility andutilization. This translates to morebang for the buck.This e-book is a beginners’ guide

to virtualization with chapters ondisaster recovery (DR), test anddevelopment, server consolidation,high availability, storage, security andkey technologies including VMwareESX, Microsoft Hyper-V and CitrixXenServer.Virtualization simplifies DR. A

physical or virtual server working inthe primary data center can be copiedinto a virtual instance on almost anyother server hardware located at theDR site. The DR site can run the virtu-al instance regardless of differencesin the remote server's hardware.We’ll explain how all this works inthe chapter on DR.Server consolidation and test and

development are the meat and pota-toes of virtualization. When develop-ers were frustrated that they couldn’t

obtain the hardware needed to testtheir software, VMware was born.This is still where most IT depart-ments start out with virtualization,and where much of the gains canbe realized.Heartened by the ability to run

multiple machines on a single server,administrators often skip the effectsto security and storage infrastructure.Fixing these problems after the fact isa headache.The security chapter explains best

practices for securing virtual machineguests on a single server as well asmeeting various compliance issues.The last chapter of this e-book

covers three main server virtualiza-tion technologies—VMware ESX,Microsoft Hyper-V and the opensource platform, Citrix XenServer,as well as the pros and cons of each.Each chapter gives you all you need

to know to enter the virtualizationgame. Once you’ve downloaded thisentire e-book, you’ll be ready for ourAdvanced Enterprise Virtualizationseminars.

Jo MaitlandExecutive Editor

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

2 INTRODUCTION TO VIRTUALIZATION

EDITOR’S NOTE:

How virtualization changes the game

Virtualization concepts in

disasterrecovery

irtualization technol-ogy is changing theface of disaster recov-ery (DR). TraditionalDR-based backup andrestoration techniques

typically imposed high IT costs asadministrators anguished over deli-cate hardware dependencies—oftendriving some companies to ignore oroverstate their DR preparedness.While virtualization itself adds

complexity to the environment, it alsobrings consolidation and flexibilitythat allows organizations to copy livestorage and server workloadsbetween remote sites running radi-cally different hardware.Virtualization is changing DR and,

with that, altering the way that datacenter administrators look at basichardware and software considera-tions, testing and performance. Con-cerns for regulatory compliance andsecurity may also enter into theequation as virtualization continuesto evolve.

VIRTUALIZATION’S EFFECTON DISASTER RECOVERYIt might seem like a simple matterto back up a workload and restoreit to another machine, but there prac-tical concerns that further complicatethis objective. Central issues are dis-ruption, hardware dependencies andcost implications of both. First,migrating copies of workloadsbetween physical hardware is disrup-tive, limits application and databaseavailability and can potentially resultin lost productivity or sales.In addition, workloads and storage

volumes are notoriously sensitiveto physical hardware—the slightestdifference could easily cause opera-tional problems with workloads atthe warm/hot remote site, as wellas problems restoring them to theprimary site.Errors can dramatically extend

recovery beyond the anticipatedrecovery time objective (RTO) andstress out an already overworked ITstaff. In response, companies typi-

CHAPTER1 Virtualization concepts in disaster recovery

3 INTRODUCTION TO VIRTUALIZATION

Va

EDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

4 INTRODUCTION TO VIRTUALIZATION

cally absorbed the cost of duplicatehardware and upgrade expenses.Some organizations have tried to mit-igate this cost by omitting non-criticalworkloads from the DR plan, relyinginstead on the availability of commonbackups for later restoration.Although virtualization seems to

be mostly beneficial, no technologyis perfect. Data center professionalsmust consider all sides in their DRplan.

q The pros: There are several bene-fits to virtualization, the most impor-tant of which is hardware independ-ence. Virtualization creates a layerof software abstraction—the virtualoperating system or hypervisor—between the workload and its under-lying hardware. This decouples thetwo and allows workloads to reside oroperate in a virtual instance unrelatedto the server and storage hardwareunderneath. Now a physical or virtualserver working in the primary datacenter can be copied into a virtual

instance on almost any other serverhardware located at the DR site. Thissite can run the virtual instance with-out concern for differences in theremote server’s hardware. It then canbe restored to the original physicalserver or migrated to another virtualserver.Such flexibility enhances server

availability. When a fault occurs onone virtual server, workloads can bemigrated to another available server.Server virtualization also enableshardware consolidation, allowingmultiple virtual instances to operateon a properly configured server.A company can use different

servers in the DR site and can man-age with fewer physical servers ifeach one is running multiple virtualinstances. The benefit of reducedhardware becomes more obviouswhen the primary site is also virtual-ized. There will be fewer servers atthe primary site and virtual workloadscan be moved non-disruptively. Syn-chronization can occur between sites

What is Recovery Time Objective?THE RECOVERY TIME objective (RTO) is themaximum tolerable length of time that a com-puter, system, network or application can be down after a failure or disaster occurs.The RTO is a function of the extent to which the interruption disturbs normal opera-tions and the amount of revenue lost per unit time as a result of the disaster. Thesefactors in turn depend on the affected equipment and application(s). The RTO ismeasured in seconds, minutes, hours or days, and is an important consideration indisaster recovery planning (DRP). —WHATIS.COM

without affecting applications anddata availability.“Storage virtualization can be

included in a virtual DR environment,but it’s not a requirement for DR,”said Ray Lucchesi, president and

founder of Silverton Consulting, anindependent technology consultantheadquartered in Broomfield, Colo.Lucchesi said that many enterprise-class network-attached storage(NAS) and Fibre Channel (FC)

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

5 INTRODUCTION TO VIRTUALIZATION

Minding your Ps and VsAS VIRTUALIZATION TECHNOLOGY garners greater attention in disaster recovery(DR) plans, you must understand the differences between physical and virtualdata migrations.

qWith a physical-to-virtual (P2V) approach, physical servers or storagevolumes at the primary site are replicated to VMs or virtual storage volumesat the DR site. This includes migrating operating systems, applications anddata.P2V is used most often in organizations that want to reduce the server or

storage system count at the secondary site, allowing complete replication ofthe primary environment onto radically different hardware. However, restoringVMs or storage volumes back to physical counterparts can be problematic ifany hardware changes take place at the primary site.

q Virtual-to-virtual (V2V) is the most flexible approach, allowing VMs andstorage volumes to be replicated between (and within) the primary data cen-ter and DR site, regardless of the underlying physical hardware involved. V2VDR supports operating systems, applications and data; recovery is simplifiedbecause it’s not necessary to meet specific hardware requirements. Tools likeVMware's VMotion can automate the process.

q In a virtual-to-physical (V2P) environment, the OS, applications and dataof a VM or storage volume are migrated to a physical counterpart at the DRsite. V2P may appear in any number of backup scenarios, such as restoringa tape backup or virtual entity, to a physical server or disk array. It may bedifficult to run the virtual workload on the physical machine, unless that physi-cal server or storage systemmeets minimum hardware and configurationrequirements. �

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

storage subsystems natively supportdata replication.Storage virtualization certainly

improves flexibility in the creation,control and non-disruptive migrationof storage volumes, but it doesn’toffer the consolidation found in servervirtualization. One TB of storage atthe primary site still requires 1 TB ofstorage at the DR site.The collective benefits of virtual-

ization bring cost savings to an organ-ization. There are fewer physical sys-tems to contend with, reduced serverhardware costs, reduced switch portcounts, less wiring and floor spaceas well as lower power and coolingdemands. Virtualization allows youto reallocate older servers or storagesystems from a technology refresh atthe primary site to the DR site, savingadditional hardware capital expense.Virtualization’s hardware-agnostic

and flexible nature eliminates mostrecovery problems and aids availabili-ty, minimizing downtime and the bur-den on IT staff. Hardware savingseasily make up for the cost to licenseand maintain virtualization software.

q The cons:Although there are fewdisadvantages to virtualization, youstill must consider them in your DRplanning. The abstraction layer need-ed for virtualization will add somecomplexity to the environment, whichrequires more management and maycomplicate troubleshooting since itcan be difficult to track virtual prob-

lems to a physical source.Consolidation is an even bigger

issue; a fault on one physical serverwill affect all virtual machines (VMs)running on that server. High-availabil-ity architectures such as server clus-tering may be needed to guardagainst hardware faults.Servers hosting multiple VMs must

be configured with CPU, RAM, I/Oand network resources adequateenough to support expected utiliza-tion. Otherwise, all VM performanceon the server will suffer. This may notbe a serious concern for a DR site thatis simply used for restoration. “They[administrators] could potentially getby with less processing power at thesecondary site and less operationalcosts,” Lucchesi said. “But they spenda bit more on [virtualization] soft-ware to support that.”

SUPPORTING VIRTUALIZATIONIN DISASTER RECOVERYSuccessful DR doesn’t happen byaccident. Actual infrastructurerequirements can vary dramaticallydepending on DR goals and corporatebudgets. Success requires carefulconsideration of several factors: net-work, server, storage and softwareresource considerations.

q LAN/WAN resources:An infra-structure evaluation normally beginswith network resources (LAN/WAN)that connect primary and secondary

6 INTRODUCTION TO VIRTUALIZATION

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

7 INTRODUCTION TO VIRTUALIZATION

sites. RTO influences bandwidthneeds.Synchronous replication duplicates

each write between sites and main-tains real-time site synchronization.However, this bandwidth-intensiveapproach generates the highest costsand is limited in distance. Asynchro-nous replication provides periodicsynchronization between sites astime permits. This can achieveacceptable results over greater dis-tances using lower-cost WAN con-nections like T1 (or slower), thoughmore time must be allocated forthe recovery point objective (RPO).When evaluating bandwidth,

remember that replication processesshouldn’t impair user access or otherdaily network performance. “If you’redoing several writes at the primarysite, the workload across the networkincreases,” Luchessi said. If the DRsite will run ”hot“ to support actualusers, there should be enough band-width and network connectivity tosupport anticipated user load.Fibre Channel over Ethernet

(FCoE), which is designed to supportIP and storage traffic on the sameinterface, can also influence virtual-ization. Not only does FCoE promiseto simplify storage networking, it canalso reduce interface costs throughthe use of integrated adapters suchas the Emulex LP21000 series.“We’re talking about combining a

fiber HBA and a network card in oneinterface at the server and the switch

level,” said Pierre Dorion, data centerpractice director at Long View Sys-tems, an IT solutions and servicescompany headquartered in Denver.Dorion pointed out that using fewernetwork adapters simplifies serverhardware requirements and reducespower consumption. He added thatnetwork virtualization can furtherenhance the flexibility of a virtualizeddata center or DR site.

q Server size and performance:While there are no definitive require-ments for a virtual server, you mustsize the server’s resources—CPU,memory, I/O, and network connectiv-ity—according to the number of virtu-al workloads that will reside on it.Experts recommend testing a VMsetup prior to an actual rollout toensure that application performanceremains acceptable on the targethardware. Once this is determined,administrators can recommendhardware upgrades or test variousconfiguration or resource throttlingoptions to ensure that each VMruns properly.Backups can hinder virtual

server performance, so evaluate themachine’s application performancewhile replicating VMs in a DR setting.”Backup is probably the most I/O-intensive operation you can subject aserver to,” Dorion said. ”If you’re run-ning 15 servers at a time on one phys-ical system, you can really bring a

(Continued on page 9)

IGH POWER AND coolingcosts convinced Pruden-tial Fox & Roach Realtors,based in Devon, Pa., toconsolidate and virtualize

half of its server infrastructure and, inthe process, simplify its DR operation.The firm replaced nearly 100 HP

servers with an IBM BladeCenter systemrunning VMware ESX virtualization soft-ware. Web servers, intranet servers,custom real estate applications, file andprint servers and monitoring systemsnow run on virtual machines (VMs).Withsubstantially fewer physical servers, thecompany expects to save $60,000 ayear on power and cooling costs.Prudential Fox & Roach Realtors runs

four BladeCenter H chassis in its centraldata center as well as one BladeCenter Schassis in a separate location to supportall of its remote offices. This secondarysite acts as a DR facility.The company replaced SunMicrosys-

tems’ tape storage equipment with IBMN-Series NAS to securely and centrallymanage company assets and clientdata. Prudential takes snapshots of allits VMs and stores them on the N-Seriesat the secondary site.“By bringing backup online and get-

ting rid of tape, we can rebuild machinesin minutes instead of days,” said Tory

Skyers, IT consultant for Prudential. Pre-viously, the firm had to rebuild the oper-ating systemmanually and would needa person in the secondary data center toinstall the backup client, perform therestore from tape or disk and reconnectthe new server to the shared storage.Now, recovery at the DR site involves

creating a new VM and redeploying thevirtual machine disk format (VMDK)over the network to this machine, whilerestoring the data from the snapshot.The VMDK file contains the whole oper-ating system environment to enablesimple loading and saving of the VM.Not only does this file contain the imageof the operating system and applicationcode, but it also describes the requiredconfiguration of the VM, including virtu-al processors, memory and devices.This serves as a simple and portable

single file that contains everythingneeded to describe the server environ-ment and the actual code and data thatmakes up that server.The new DR strategy was put to the

test when a virus brought down one ofthe firm’s servers. A snapshot of theserver existed at the secondary site, andit was brought back online within twohours. “We can conduct business if onedata center goes down without breakingthe bank,” Skyers said. —J.M.

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

8 INTRODUCTION TO VIRTUALIZATION

CASE STUDY:

Prudential easesDRwith virtualization

H

physical server to its knees.”The performance implications

of backup/recovery processes canradically affect consolidation choicesfor an enterprise. For example, fiveVMs might run well on one physicalserver, but it will slow to an unaccept-able crawl during a backup. Dorionnoted that there’s nothing wrong withhosting a single VM such as a SQLServer on a single physical machine,which still produces non-disruptivemigration.

q Effects on storage: Virtualizationadds flexibility and non-disruptivemigration to storage, though virtualstorage is not an essential elementof virtual DR. There are no minimumstorage virtualization requirements.The DR site will require enough

total disk capacity to hold storagevolumes migrated from the primary

site, but that storage does not haveto be virtualized. Servers must havelocal storage large enough to retainall of the VM images intended forthat physical system, so the DRservers and their local storage arealmost always virtualized.

q Third-party software tools: Soft-ware tools are essential for replicat-ing VMs and storage volumes; manyDR tools are available from the virtu-alization vendor. Citrix XenServer5has built-in live migration and DRfeatures. Other vendors like VM-ware provide separate free toolslike VMware Consolidated Backup,which is designed to protect an entireVMware environment under VMwareInfrastructure 3. VMware VMotioncan move VMs non-disruptivelybetween physical servers; evenWin-dows Server 2008 R2 includes livemigration of VMs between servers.

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

9 INTRODUCTION TO VIRTUALIZATION

Are you ready for a DR SAN?A VIRTUALIZED DR site is best served with a storage area network (SAN). Thecentralized high-performance storage provided by a SAN adds redundancy,rather than relying on individual disks within physical servers. It also enablesthe rapid migration of virtual images between systems. Cost-conscious organ-izations, though, may be reluctant to fund a conventional FC SAN for a remoteDR site. Instead, organizations may consider an iSCSI SAN deployment to sup-port storage and IP traffic on the same network. For many organizations, iSCSIrunning on a 1 GB Ethernet network strikes an acceptable balance betweenaffordability and performance. �

(Continued from page 7)

Server clustering software canalso play a role in a virtualized datacenter or DR environment; VeritasCluster Server from Symantec Corp.is just one example. Clusteringenables physical servers to work inconjunction with each other, sharingtheir collective computing resourcesto handle more workloads or improveperformance. Clustering also sup-ports high availability—if one physicalserver fails, the remaining clusterelements take control and continueprocessing.Other third-party software, such

as vFoglight from Vizioncore Inc., canfurther enhance a virtualized environ-ment by monitoring VM perform-ance. Vizioncore’s vReplicator canalso enhance DR and high availabilitywithin a VMware environment.

TESTING AND IMPROVINGPERFORMANCEOne of the principle weaknesseswith DR is inadequate testing. Tradi-tional DR required that recoveryoccur to a working environment. Thisresulted in unacceptable disruptionsat the primary site, especially if recov-ery errors happened. The alternativewas to attempt recovery to a test site,but recreating and maintaining addi-tional redundant recovery hardwareproved cost-prohibitive. When virtu-alization is used to consolidateservers in the DR site, it’s simpler totest the recovery of many VMs to a

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

10 INTRODUCTION TO VIRTUALIZATION

VIRTUAL DR CHECKLIST

Test DR regularly: Virtualizationsimplifies the DR-testingprocess. Take every opportunityto test and verify disaster readi-ness.

Don’t ignore upgrades: Hypervi-sors and other software residingon servers at the DR site must beincluded in regular patch/updatecycles.

Choose servers carefully: Selectservers with computing power(CPU, memory, I/O and networkconnectivity) adequate enoughto support the intended numberof VMs and users.

Back up the backup: ProtectVMs at the DR site using a SAN,or back up VMs using otherstorage like NAS or tape.

Balance virtualization withperformance: Rather than backup multiple VMs simultaneously,stagger the process or reduce thenumber of VMs to manage theadded I/O strain.

Look beyond servers: Look foropportunities to leverage othertechnologies such as networkvirtualization or desktop virtual-ization. �

few extra virtual servers.With an increasing number of “hot”

DR sites, the availability of applica-tions and data at the site itself isoften taken as proof of DR readiness.For example, administrators can sim-ulate a disaster at the primary site by

rerouting DNS to the DR facility, andsimply allowing users to work on theSAN and virtual servers there.If that works, DR site content is

replicated back to the productionsite (often non-disruptively), andthe DNS is rerouted back to the maindata center. DR in a virtualized envi-ronment can be tested more fre-quently with less effect on users.Administrators usually don’t needto be at the DR site.Regardless of the specific approach,

testing a virtualized DR environmentshould extend beyond simply verify-ing recovery to include a performanceevaluation. “Make sure that your DRenvironment can carry the processingload,” Dorion said. “You may accept[your servers] running at a somewhatdegraded level of service during arecovery effort.”Network monitoring tools can

measure performance at the DR site,though actual end users can also pro-vide solid feedback on performancesnafus during a backup/restoration.The proper use of storage in a virtu-

al environment can improve DR. Dori-on points out that disk failure cancripple a physical DR server and leavenumerous VMs inaccessible. It’s criti-cal to protect VMs at the DR site withmore traditional backups like tape.“Virtualizing servers is no guaran-

tee of reliability,” Dorion said. “Youstill need to make sure that your datais adequately protected.” Considerstoring VM images on protected diskstorage, such as RAID, in SAN or NASarrays as an alternative. Centralizingthe storage of VM images also allowsmore efficient reallocation from onephysical server to another.

.THE EVOLUTION OF VIRTUAL DRConcerns about compliance andbusiness continuance are drivingthe need for DR, but the landscape isstill evolving. In addition to FCoE andnetwork virtualization, ever-growingcomputing power and attention tosecurity will influence the future ofvirtual DR.The number of available processors

limits server virtualization. Serversbased on inexpensive two-coreprocessors can easily migrate to morepowerful four- or six-core processorslike Intel’s Xeon 7000 series orAMD’s Opteron family. This is espe-

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

11 INTRODUCTION TO VIRTUALIZATION

Disk failure cancripple a physical DRserver and leave manyVMs inaccessible.

cially true for processors with “virtu-alization acceleration” technology,such as Intel VT or AMD-V capability.Eight-core platforms that use twofour-core processors are cropping up,and “true” eight-core processors suchas Sun Microsystems’ UltraSPARC T2will become more commonplace inthe next 18 to 24 months.Security will also play a vital role in

future virtualization. Traditional secu-rity techniques generally monitor net-work traffic and its behavior. This issuitable with distinct physical serversand networking hardware, but whenmultiple servers are hosted on thesame machine—along with networkvirtualization technologies like “softswitches”—virtual security mustemphasize interprocess monitoringof VM interaction.

Virtualization security flaws alsosurface in server configuration andOS patching. It’s much easier tooverlook a configuration setting orOS patch level when there are dozensof VMs on a physical server. Hyper-visors can also suffer from securityflaws, potentially exposing the VMsthat run under them. Virtualizationusers should add hypervisor patch-ing/updating to their OS mainten-ance process.The introduction of virtualization

has dramatically changed the tradi-tional data center environment. Asthis technology continues to evolve,so will virtual DR considerations andplanning. Having a solid plan in placeand examining each aspect will makeall the difference. �

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

CHAPTER1 Virtualization concepts in disaster recovery

12 INTRODUCTION TO VIRTUALIZATION

Cathy Gagne, Editorial Director, cgagne@techtarget.com; Christine Casatelli, Managing Editor,ccasatelli@techtarget.com;Michelle Boisvert, Features Editor, mboisvert@techtarget.com;MarthaMoore, Copy Editor, mmoore@techtarget.com; Linda Koury, Art Director of Digital Content,lkoury@techtarget.com;Marc Laplante, Publisher, mlaplante@techtarget.com©2009 TECHTARGET. ALL RIGHTS RESERVED.

ABOUT THE AUTHOR

Stephen J. Bigelow, Senior Features Writer, has more than 15 years of technical writingexperience in the PC/technology industry. He has authored hundreds of articles and morethan 15 feature books on computer troubleshooting, including Bigelow’s PC Hardware DeskReference and Bigelow’s PC Hardware Annoyances. Contact him at sbigelow@techtarget.com.

q Virtualization 101

qUsing virtualization for testing and development environments

qWhat is Microsoft Hyper-V?

About Dell Inc.:Dell Inc. (NASDAQ: DELL) listens to customers and delivers innovativetechnology and services they trust and value. Uniquely enabled by its directbusiness model, Dell is a leading global systems and services company andNo. 34 on the Fortune 500. For more information, visit www.dell.com, orto communicate directly with Dell via a variety of online channels, go towww.dell.com/conversations. To get Dell news direct, visit www.dell.com/RSS.

aEDITOR’S NOTE

aVIRTUALIZATIONCONCEPTS IN DR

aPROS AND CONS

aINFRASTRUCTURE

aCASE STUDY

aTESTING AND

PERFORMANCE

RESOURCES FROM OUR SPONSOR

13 INTRODUCTION TO VIRTUALIZATION