View Roadshow

VMware View Road Show

Agenda Trace3 Intro Define Your X NetApp Global Dynamic Lab Whats New with vSphere 4.1? Why NetApp? Whats New with View 4.5? Wyse Thin Clients Best Practices for VMware View page 2 | 2010 Trace3. Proprietary and Confidential.

Who is Trace3?Trace3 is a Systems Integrator focused on meeting the needs of companies looking for proven solutions in the following areas:

Storage Networking Data Protection Virtualization Security Microsoft

page 3 | 2010 Trace3. Proprietary and Confidential.

Trace3 LocationsSeattle

Portland

Boise Cheyenne Omaha

San Francisco

Reno

Salt Lake Las Vegas

3:1 Tech to Sales ratioDenver

Trace3 Regions

Los Angeles Phoenix Orange County San DiegoAlbuquerque

LA OC SD

Rocky Mountain Desert Phoenix Pacific Northwest

page 4 |

2010 Trace3. Proprietary and Confidential.

Some of Trace3 Customers

page 5 |


Trace3 Services by Manufacturer

page 6 |


Define Your

page 7 |


Where do you start!1 . Define Your X 2. Discover Starting point 3 . DefineRoute and time - line

5 . Deploy & Document Get

moving , get to work

6 . Validate & SupportManifest of the journey

Limit Distractions , Stay on Track , Reach the Goal6 . Validate & Support

5 . Deploy & 4 . Design 4 . Design Document Products , tools , bestpractices

page 8 |


Whats your X ?

Whats your ? How do you plan to get there?

page 9 |


Defining your X Understand your Business Requirements Build a Business Case Dont focus on Cost or Technology Define Success

page 10 |


Business Reasons for Desktop Virtualization Rein in the costs of PC Management Secure devices and data regardless of location Remain compliant with regulations and mandates Support a changing workforce Planning for disaster or workforce disruption

page 11 |


Sample Business CaseCost Manageab ility Security Mobility Reduce CapEx cost compared to traditional PC Reduce deployment times for desktops Reduce OpEx cost associated with PC Centralize management Meet compliance requirements via Management storage Reduce recovery times for PC centralized Provide constant work environment in Extend desktop refresh cycle to 5 - 6 failures Deploy comprehensive security the office or on the road years framework Enable Bring Your Own PC ( BYOPC )

page 12 |


Your X DefinedSecure Anytime Anyplace Any device Automated Provisioning Private Cloud

VMware vSphere

VMware vSphere

page 13 |


Servers?! Intel Core 2 Duo Think Xeon 7500/5500

Large Rackmount Think Blade Chassis

8-16 GB RAM Think 32GB+ of RAM

L I A F

page 14 |


Network?! 100mb Hubs? Think 10GB Ethernet

Individual Interfaces Think VLAN Trunking

L I A F

page 15 |


Storage?! Direct Attached Disk Think NAS/SAN

FC vs iSCSI or NFS Think NFS or iSCSI

L I A F

Best Practices for running VMware vSphere on Network Attached Storage http://vmware.com/files/pdf/techpaper/VMware-NFS-BestPractices-WP-EN.pdfpage 16 | 2010 Trace3. Proprietary and Confidential.

Desktops High End Graphics Keep Standalone

CPU Intensive Keep Standalone

L I A F

page 17 |


Cost!? VDI is cheaper than Traditional Desktops Think Again

VDI leverages existing Virtualization Investment Think about Scale

page 18 |


NetApp Global Dynamic LabBuilding a 50,000 Seat VMware View 4.5 Deployment: A Collaboration by Cisco, VMware, NetApp, Fujitsu and WYSE

page 19 |


Agenda Project Overview Partner Collaboration The Pod Based Architecture Solution Components Benefits of shared storage in VDI Value to customer

page 20 |


50,000 Seat VDI Deployment Worlds largest documented VDI deployment Industry leaders collaborating together Cutting edge software: View 4.5, Windows 7 Validates performance and scalability of solution Proof-point for technology, integration, and partnerships

page 21 |


Goals of the Reference Architecture1.Create a POD" design that is

Scalable Cost Effective / Efficient High Performing Proven

2.Thorough solution component testing

Hypervisor, Server, Network, Storage, Thin Clients View 4.5, Windows 7 VMware RAWC tool

page 22 |


3.Provide a blueprint to customers

5000 Seat Pod ArchitectureWYSE Zero Client

VMware View 4.5 Windows 7 Vmware vSphere 4.1

Connection Broker

Hypervisor

Cisco UCS 5100 Blade Server

Cisco UCS 6100 Fabric Interconnect

Cisco Nexus 5020 Cisco Nexus 7018

NetApp FAS 3170 Cluster

page 23 |


Management ArchitectureVMware View 4.5 3 Connection Brokers (5000 Desktops) VMware View 4.5 VMware View 4.5

HA / DRS Cluster

HA / DRS Cluster

HA / DRS Cluster

HA / DRS Cluster

2500 VMs / vCenter (30 ESX Hosts)

2500 VMs / vCenter (30 ESX Hosts)

60 ESX Servers (Cisco UCS)

page 24 |


Benefits of Shared Storage for VDI Manageability Automatic Migrations DRS Hardware Assisted VM Cloning Centralized Management of Infrastructure & Images

Multi-Protocol Datastores & User Data Resiliency VMware HA Backup and recovery/data protection Disaster Recoverypage 25 | 2010 Trace3. Proprietary and Confidential.

W h a t N e tA p p T e ch n o l g y w a s u se d o M a n a g e m e n t & D e p l ym e n t o V S C 2 .0

P e rfo rm a n ce D e d u p l ca ti n i o Fl sh C a ch e a T ra n sp a re n t S to ra g e C a ch e S h a ri g n

C o st E ffe cti n e ss ve D e d u p l ca ti n & S to ra g e E ffi e n t C l n i g i o ci o n

page 26 |


A ch i ve si n i ca n t co st re d u cti n s e g fi o

Users @ 25GB Number of NetApp Gold Datastores Number of Datastores Traditional Traditional Storage Requirement NetApp Storage Needed NetApp Storage Savings NetApp Storage Savings

250 1 1 6 TB 74 GB 6 TB 99 %

5000 2 20 115 TB 148 GB 115 TB 99 %

50000 20 200 1 , 150 TB 2 , 960 GB 1 , 147 TB 99 %

Note: Data savings based on initial deployment of virtual machinespage 27 | 2010 Trace3. Proprietary and Confidential.

Pod-based architecture Easily expandable Predictable performance

Virtual Desktopspage 28 | 2010 Trace3. Proprietary and Confidential.

50 , 000

Summary Proof-point for integrated technologies Example of collaborative partnerships Validation of multi-vendor solution Reference architecture as starting point

page 29 |


NetApp Video

page 30 |


Whats new with vSphere 4.1?

page 31 |


VMware Business Infrastructure Solution AreasDatacenter SolutionsConsolidationDecrease capital expense Simplify infrastructure Delay facility expansion

Enabling a more dynamic, flexible service model, at lower costs

ManagementDecrease operating expense Manage SLAs , not components Automate the mundane , simplify the complex

Business ContinuityEnsure high availability at low cost Recover quickly and efficiently Guarantee service levels App Dev

Desktop SolutionsEnabling desktop as a serviceUser access across boundaries Desktop as a service , without the headache Increase security and control

Cloud ServicesEnabling the bridge to the external cloud

Optimizing application development to deploymentSimplify test set - up and execution Speed time to market and improve quality Leverage modern application framework

& Test

Single way to run , manage and secure applications both internal and via external clouds Broadest choice of leading cloud service providers

VMware vSphere

page 32 |


VMware vSphere 4.1: Whats New?vCenter Server ( 64 - bit )HA Diagnostic s And Healthcheck vMotion Speed and Scale Update Manager Enhancements Virtual Serial Port Concentrator More VMs ( per cluster , DC ) More Hosts ( per VC , DC )

s c i v r e S

n o t a c i l p A

AD Integratio n (host)

Availabi lity

Security

Scalabil ity

VMware vSphere 4.1s c i v r e S e c u t s a r f n IvCompute vStorageStorage I/O Control More Performanc e Metrics APIs for Array Integration

vNetworkNetwork I/O Control Load Based Teaming IPv6 NIST Performance +

Memory Compress ion Host Affinity

page 33 |


vSphere 4.1 Delivers Cloud Scale

3,000 VMs / cluster (2x) 500 hosts / vCenter (5x) 10,000 VMs / vCenter (3x)

99% of VMwares 170K Customers Can Run Their Entire Datacenter in a Single VMware Cluster*page 34 | 2010 Trace3. Proprietary and Confidential.

E n h a n ce d S ca l b i i D e fi e d a l ty nvSphere 4 vSphere 4 . 1 RatioVMs per host Hosts per cluster VMs per cluster Hosts per VC

320 32 1280 300

320 32 3000 1000 15000 10000 120 500 5000

1x 1x >2x >3x >3x >3x 4x 5x 2x

Registered VMs per VC 4500 Powered - On VMs per VC 3000 Concurrent VI Clients 30 Hosts per DC VMs per DC

100 2500

page 35 |


Migration to ESXi with vSphere 4.1 Recommended that all vSphere 4.1 deployments use the ESXi Hypervisor vSphere ESXi 4.1 Fully Supports Boot From SAN for FC, iSCSI and FCoE vSphere 4.1 is the last release with the ESX hypervisor (ESX with Service Console) Visit the ESXi Upgrade Center -

http://www.vmware.com/products/vsphere/esxi-upg /

page 36 | 2010 Trace3. Proprietary and Confidential.

ESXi to ESX Info Center

page 37 |


vCenter Server Migration to 64-bit vCenter Server MUST be hosted on 64-bit Windows OS 32-bit OS NOT supported as a host OS with vCenter vSphere 4.1

Why the change? Scalability is restricted by the x86 32 bit virtual address space and moving to 64 bit will eliminate this problem Reduces Dev and QA cycles and resources (faster time to market)

Two Options1. vCenter Server in a virtual machine running 64-bit Windows OSpage 38 | 2010 Trace3. Proprietary and Confidential.

Storage I/O ControlDescriptionSet storage quality of service priorities per virtual machine

BenefitsBasic - Make storage access rights equal between VMs Advanced - Prioritize use of storage (similar to how compute is prioritized with vSphere) per VM Business priorities now define low and high priority storage resource access Create the high speed or HOV lane for VMs

Beta Feedback

I really feel that the Storage I/O Control is a must have for our environment and we should move forward without delay.

Proof Point

1 . All VMs created equal Guarantee service levels for access to storage resourcespage 39 | 2010 Trace3. Proprietary and Confidential.

2 . Make Your Mission Critical VMs VIPs

Storage I/O Control (SIOC)CPU shares: Highonline store

CPU shares: HighMicrosoft Exchange

CPU shares: Lowdata mining

Memory shares: High

Memory shares: High

Memory shares: Low

I/O shares: High

I/O shares: High

I/O shares: Low

32GHz 16GB

Datastore A

page 40 |


Storage Performance ReportingDescriptionDeliver of key storage performance statistics in vCenter

BenefitsGranular storage reporting for improved tuning and troubleshooting performance Independent of storage architectures and protocols

Beta Feedback

In the monitoring area, the enhanced storage statistics are very useful

Proof Point

Real-Time and Historical Trending for Storage

page 41 |


Network I/O ControlDescriptionSet network quality of service priorities per flow type (iSCSI, NFS, etc.)

BenefitsBasic - Make network access rights equal between flow types Advanced - Prioritized use of network, especially in 10 Gbit environments Business priorities now define low and high priority network resource access as needed Create the high speed or HOV lane for VMs iSCSI FT NFS

Beta Feedback

The new Network I/O control feature is very interesting for consolidating network links with 10Gbit.

vMotion

Proof Point

TCP/IP

Guarantee service levels for access to network resourcespage 42 | 2010 Trace3. Proprietary and Confidential.

Distributed Switch

10 GigE

vMotion Performance and Scale EnhancementsDescriptionAdding Cloud Scale to online virtual machine migration (a VMware key differentiator) Beta Feedback This release product has some nice benefits in particular increased vMotion capabilities.

BenefitsPerformance and Scalability More Live Migrations in Parallel (up to 8 per host pair) Elapsed time reduced by >4.5x on 10GbE tests

Point

Proof

5x faster with the 4.1 platform release

page 43 |


Memory CompressionDescriptionA new hierarchy for VMwares memory overcommit technology (a VMware key differentiator)

BenefitsOptimized use of memory Safeguard for using memory overcommit feature with confidence Performance

Beta Feedback

Great for memory oversubscription.

Proof Point OS

1,000x faster than swap-in!Hypervisorpage 44 | 2010 Trace3. Proprietary and Confidential.

DRS Host AffinityDescriptionSet granular policies that define only certain virtual machine Movements

BenefitsTune environment according to availability, performance, and/or licensing requirements Cloud enablement

Beta Feedback

Awesome, we can separate VMs between data centers or blade enclosures with DRS host affinity rules

Proof Point

Mandatory Compliance Enforcement for Virtual MachinesVMs A Servers A Onlypage 45 | 2010 Trace3. Proprietary and Confidential.

VMs B Servers B Only

HA EnhancementsDescriptionHealthcheck status Operational window Optimized interaction with DRS Application-Aware API Beta Feedback Major improvements in DRS!

BenefitsEvent or alarms when configuration rules are broken No click status (cluster status available at all times) Move VMs to the Best Host Available Application awareness (with supported solution)

Point

Proof

Adding Another 9 to Availability

page 46 |


Fault Tolerance (FT) EnhancementsDRS

FT Primary VM

FT Seconda ry VM

FT fully integrated with DRSDRS load balances FT Primary andSecondary VMs. EVC required.

Versioning control lifts requirement on ESX build consistencyResource Pool

Primary VM can run on host with adifferent build # as Secondary VM.

Events for Primary VM vs . Secondary VM differentiatedEvents logged/stored differently.

page 47 |


vStorage APIs for Array Integration (VAAI)vStorage APIs for Array Integration

VMware vSphere

Storage vMotion

Provision VMs From Template

Improve Thin Provisioning Disk Performance VMFS Share Storage Pool Scalability

page 48 |


Storage vMotion with Array Full Copy Function Benefits Zero-downtime migration Eases array maintenance, tiering, load balancing, upgrades, space mgmt

Challenges Performance impact on host, array, network Long migration time (0.5 - 2.5 hrs for 100GB VM) Best practice: use infrequently

Improved solution Use arrays native copy/clone functionalitypage 49 | 2010 Trace3. Proprietary and Confidential.

VAAI Speeds Up Storage vMotion - Example

42:27 - 39:12 = 2 Min 21 sec w/out (141 seconds) 33:04 - 32:37 = 27 Sec with VAAI 141 sec vs. 27 sec

page 50 |


VM Provisioning from Template with Full Copy Benefits Reduce installation time Standardize to ensure efficient management, protection & control

Challenges Requires a full data copy 100 GB template (10 GB to copy): 5-20 minutes FT requires additional zeroing of blocks

Improved Solution Use arrays native copy/clone & zeroing functionspage 51 | 2010 Trace3. Proprietary and Confidential.

Up to 10-20x speedup in provisioning time

C o p yi g D a ta O p ti i d C l n i g w i n m ze o n th VAAIVMFS directs storage to move data directly Much less time!Up to 95% reduction

Dramatic reduction in load on:Servers Network Storage

page 52 |


Scalable Lock Management A number of VMFS operations cause the LUN to temporarily become locked for exclusive write use by one of the ESX nodes, including: Moving a VM with vMotion Creating a new VM or deploying a VM from a template Powering a VM on or off Creating a template Creating or deleting a file, including snapshots

A new VAAI feature, Hardware Assisted Locking (atomic test and set) allows the vSphere host to offload the management of the required locks to the storage and avoids locking the entire VMFS file system. 2010 Trace3. Proprietary and Confidential.

page 53 |

V M FS S ca l b i i w i H a rd w a re A ssi d a l ty th ste Lo cki g n M a ke s V M FS m o re sca l b l o ve ra l , b y a e l o ffl a d i g b l ck l cki g m e ch a n i o n o o n sm

Using Atomic Test and Set (ATS) capability provides an alternate option to use of SCSI reservations to protect the VMFS metadata from being written to by two separate ESX Servers at one time.

Normal VMware Locking (No ATS)

Enhanced VMware Locking (With ATS)

page 54 |


For more details on VAAI vSphere 4.1 Documentation also describes use of this features in the ESX Configuration Guide Chapter 9 (pages 124 - 125) Listed in TOC as Storage Hardware Acceleration Three setting under advanced settings: DataMover.HardwareAcceleratedMove DataMover.HardwareAcceleratedInit VMFS3.HarwareAcceleratedLocking Assisted Locking Full Copy Block Zeroing Hardware

page 55 |

Will only support block based storage in 4.1 NetApp Integration with VMware 2010 Trace3. Proprietary and Confidential.

Why NetApp?

page 56 |


Unified: much more than multi-protocolProtocolsFC FCoE iSCSI NFS CIFS

Low-to-High Scalability

Unified Management Same tools & processes: learn once, run everywhere Integrated data management Integrated data protection

Multi-Vendor Virtualisation

Cost/PerformanceFlash Cache SSD FlexCache

One Architecture, One O/Spage 57 | 2010 Trace3. Proprietary and Confidential.

NetApp Flash Cache

page 58 |


45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

D S4243

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

Alleviates Storms Reduces Disk IO Deduplication Aware Extend and Optimize

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

D S4243

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

D S4243

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

D S4243

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

page 59 |


Benchmarked FAS3160 Configs224 FC drives 64TB

Reduc e45 0GB 4 5 0 GB 4 5 0 GB 45 0GB 4 5 0 GB 4 5 0 GB 45 0GB 4 5 0 GB 4 5 0 GB 45 0GB 4 5 0 GB 4 5 0 GB 45 0GB 4 5 0 GB 4 5 0 GB 45 0GB 4 5 0 GB 4 5 0 GB

75% Fewer More Spindles Capacity

50%

45 0GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

Cost Rack Space Power Cooling4 5 0 GB

96 SATA drives 56 FC drives 16TB 96TB

D S4243

4 5 0 GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

D S4243

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

45 0GB

4 5 0 GB

4 5 0 GB

4 5 0 GB

FC Baseline Configuration

FC + Flash Cache SATA + Flash Cache Configuration

Fewer Drives, Equal Performancepage 60 | 2010 Trace3. Proprietary and Confidential.

page 61 |


VM DISK

VM DISK

VM DISK

Remember NetApp Does the Same with Transparent Page Storage Sharing?

NetApp Deduplication

page 62 |


Transparent Storage Cache Sharing

AND

page 63 |


ESX Cluster VM Datastore VM VM VM

Cache Deduplication Data

Set Without Extended BuiltItinto Works And with Data with Forget Migration Flash It! ONTAP ANY Cache Dataset

VM

VM

VM

VM

TSCS Storage Cache

VM

VM

VM

VM

Disk Storage NetApp FAS Array Traditional Array

page 64 |


page 65 |


Virtual Servers AND Desktops

page 66 |


Exchange 2010

PsstExchange Single Instance Storage is going away!page 67 | 2010 Trace3. Proprietary and Confidential.

Deduplicate AND Data Works with Encapsulated AccelerateOST OST OST

User Data Disk (VMDK)



VMFS or NFS Datastore NetApp WAFL (dedupe layer)

Transparent to UsersPsst Its not file level dedupe and compressionpage 68 | 2010 Trace3. Proprietary and Confidential.

NetApp VSC 2.0

page 69 |


Automated setting of best practices

page 70 |


page 71 |


page 72 |


page 73 |


page 74 |


Datastore VM VM VM VM

Virtual Machine

Cloning

VM Template Datastore

NetApp Storage Array

page 75 |


Datastore VM VM VM VM

Virtual Machine

Redeploy

VM Template Datastore

NetApp Storage Array

page 76 |


Data Protection for Any Failure

VMware HA

VMware vCenter Site Recovery Manager

MetroClu ster

Thin Replication

Building 1

Building 2

DR Site

Issue User error : Data corruption: System or site failure: Regional disasters: Planned Downtime:page 77 | 2010 Trace3. Proprietary and Confidential.

NetApp Solution End-user recovery (Snapshots ) Instantaneous restores Transparent failover Recover in minutes Zero planned downtime

Application MobilityPrimary Site Secondary Site

AP P O

A A SP S P O O P P S S

AP P O

A P O P S

A P O P S

Host load balancing or maintenance Datastore capacity management

Use Case

VMware VMotion VMware Storage VMotion

Solution

page 78 |


78


AP P O

A SP O P S

AP P O

A SP O P S

A P O P S

A P O P S

Host load balancing or maintenance Datastore capacity management Array balancing, maintenance or refreshpage 79 | 2010 Trace3. Proprietary and Confidential.

Use Case

VMware VMotion VMware Storage VMotion NetApp DataMotion

Solution

79


Application Application data availability access not Impacted over long distances using NetApp can introduce FlexCache latency

AP P O

A SP O P S

AP P O

A SP O P S

A P O P S

A P O P S

A P O P S

Host load balancing or maintenance Datastore capacity management Array balancing, maintenance or refresh Data mobility without affecting availability page 80 | 2010 Trace3. Proprietary and Confidential.

Use Case

VMware VMotion VMware Storage VMotion NetApp Data Motion NetApp, Cisco, VMware Long Distance VMotion

Solution

Dev/QA Environment

Production Environment

Desktop Environment

VMware

VMware

VMware

Secure environments traditionally everything Virtualize require dedicated resources Inefficient and inflexible Costly to deploy Unified computing and manage Low utilization rates Difficult to change Unified fabric one wire How to gain efficiencies of virtualization while maintaining security?

Unified storageDev/QA Prod Desktop

Secure Multi-tenancy

page 81 |


Multi-TenancyCloud Administrator

NetApp MultiStorevFiler vFiler vFiler vFiler

Tenant A

Tenant B

Tenant C

Tenant D

Tenant B

page 82 |


Additional References VMware View Reference Architecture VMware View http://www.vmware.com/products/view/

http://www.vmware.com/resources/techresources/10

VMware VDI Best Practices on NetApp Storage http://media.netapp.com/documents/tr-3705.pdf

VMware / Cisco / NetApp2000 Seat VDI Reference Architecture http://media.netapp.com/documents/tr-3770.pdf

Deploying VDI with VMware View and NetApp

http://www.vmware.com/files/pdf/resources/VMware_page 83 | 2010 Trace3. Proprietary and Confidential.

Whats new with View 4.5?

page 84 |


V i a l za ti n & C l u d M a n a g e m e n t rtu i o oEnd User Computing

Provision & deploy desktops rapidly Manage workstation images simply Deliver virtualized applications to desktop

End - User Computing Management

Cloud Applications

Encapsulate applications into containers IT Business Managemen with vApps Govern cloud provisioning pr Assure application portability & performance cloud software license Track Establish service contracts with infrastructure Standardize cloud partitions

Application Management

Infrastructure & Operations ManagementPublic / Private / Hybrid Cloud Virtualized Infrastructure

Create a zero-touch, compliant infrastructure Build automation into platform Deliver self-service through policy-driven control

VMware Management Solutionspage 85 | 2010 Trace3. Proprietary and Confidential.

Deliver desktops as a managed servicePlatformVMware vSphere for desktops

Management

VMware View Manager Vmware View Composer VMware ThinApp

User ExperiencePCoIP Print Multi-monitor display Multimedia USB redirection

page 86 |


Components of Desktop as a Managed ServiceUsability Flexibility

User Experience

View Client PCoIP Protocol Local Mode

Reduce IT Costs

Simplicity Efficiency Security

Management

View Manager View Composer ThinApp

Availability Reliability Scalability

Platform

vSphere for Desktops

page 87 |


PCoIP Improvement and Changes Smart Card Support & Online Certificate Status Protocol (OCSP) Certificate Revocation

Location Based Printing & Awareness Custom display topology with zero clients FIPS140-2 compliance application Port Change to 4172 Improved WAN Performance Details on next slide

page 88 |


PCoIP WAN Improvements in View 4.5 Four changes added to View 4.5 compared to improve WAN performance: Improved image quality management Improved network bandwidth estimation Improved out of order packet resilience Imaging selective packet retransmit

Anticipated Impacts on WAN Experience Higher image quality without impacting bandwidth consumption Improved imaging performance in low bandwidth situations Improved performance when sharing network with multiple PCoIP sessions or other TCP trafficpage 89 | 2010 Trace3. Proprietary and Confidential.

No service degradation when tested with Juniper, F5, OpenVPN, and Cisco SSLVPN solutions

Why PCoIP protocol does well on WAN PCoIP protocol uses host-side rendering to avoid client redirection latency limitations PCoIP protocol uses UDP to transfer realtime audio and image data optimally (no resending of stale packets required) PCoIP protocol dynamically adjusts image quality and frame rate based on available bandwidth PCoIP protocol is able to use lossy compression on images and audio when network is constrained 2010 Trace3. Proprietary and Confidential.

page 90 |

Planning for PCoIP Plan for 200-250kbps average bandwidth for a typical basic office productivity desktop Plan for 500kbps - 1 Mbps minimum peak bandwidth to provide headroom for bursts of display changes Plan for 1Mbps per simultaneous user running 480p video Plan for less than 70-80% network utilization Assumptions are based on 8-10 hours continuously usage, Bandwidth estimation (250Kbps/0.80) * 5 = 1.5Mbps T1page 91 |


Win7 OS Support Guest and Client Supporting Windows 7 32-bit and 64-bit as a client and remote desktop Virtual desktop related improvements Jumplist integration GPO PowerShell 2.0 cmdlets Location-aware printing RDP7

True multi-monitor and Multimedia redirection support MMR is enabled by default and cannot be disabledpage 92 | 2010 Trace3. Proprietary and Confidential.

Aero support for RDP7 client

Mobility & Bring Your Own Computer View Client with Local Mode (Type 2 Hypervisor) Virtual desktop is checked-out to local endpoint and encrypted Access desktop, applications and data regardless of network availability Changes are checked in to the datacenter when online Extend IT security policies to local desktopView Manager

page 93 |


W h a t s ch a n g e d si ce e xp e ri e n ta l n m O ffl n e V D I i ?CapabilityRobust Transfer Infrastructure Centralized Data Control Improved Security

DetailsSeparate "Transfer Server" instances for scalable download/upload of VM data VM data is optimized, compressed & deduplicated (e.g. no more copying page-file bits) Optional use of 3rd party HTTP proxy caches for optimized provisioning to remote sites View Composer derived base images can be pre-loaded via alternate media

Admin can initiate "replication" of user changes back to server on demand Admin policy enforcement of whether or not users can initiate checkin or replication

Disaster Recovery Infrastructure Cost Savings Latest Runtime Platform

Checked out desktops all encrypted with AES (128-bit by default, 256-bit option) SmartCard Authentication Support Admin can remotely revoke end user access to a checked out desktop* Admin can set policy for how frequently a client must get in touch with the server for updates/enforcement (or be deactivated)* Policy controls to allow/disallow access to USB devices like thumb-drives* Delta-Disk replication can be scheduled to occur automatically If client device is subsequently lost/destroyed, desktop can be easily reconstituted from last replicated image

Pools can be configured as "local only" to save on server & storage CAPEX Supports VM provisioning with View Composer Checked out VMs can be based on Linked Clones for further storage cost reduction

Uses latest VMware virtualization platform for great end user experience... High quality multimedia & audio (play HD moves, YouTube videos, etc) 3D with DirectX9c and Windows 7 Aeroglass Effects support Hardware accelerated encryption/decryption with Intel AES-NI capable chipsets Support for Windows 7, XP and Vista host and guest Operating Systems

page 94 |


Local Mode Administration All local desktops: Require user authentication to run Are completely encrypted Must periodically heartbeat with View Connection Server for management

Policies affecting local desktops: Can the desktop be used locally? How long can the local desktop go without server contact? What part of the local desktop should be replicated? (Linked clone desktops only) How often should the desktop be replicated? Is the user allowed to initiate replication? Check in? Rollback?

Rollback Discard a local desktop and make server side desktop live

Initiate Replicationpage 95 | 2010 Trace3. Proprietary and Confidential.

Schedule a one-off replication on next client contact

Smart Card Authentication Support PCoIP and RDP Revoked certificates may be published through OCSP / CRL Cached and encrypted PIN entry for Local Mode smart card logon Storage of multiple credentials for public key infrastructure (PKI), one-time password (OTP), and static passwords on a single authentication device Support for leading smart card manufacturers, remote access solutions, thin clients, and productivity applications Capability to establish specific policies for certificates, PIN management and notification Support for smart card standards direct SSO

page 96 |



User Experience

View Client PCoIP Protocol Local Mode

Reduce IT Costs


Management



Platform


page 97 |


Scalability Broker Level Broker Pod and Teaming Federated Pool Management Floating & Dedicated Pools Non-Persistent Pool Refresh & Re-Compose

View Composer Tiered Storage Local Disk Storage Support Disposable Disks All Users directory is no longer copied during customization Thin-Provisioned Diskspage 98 | 2010 Trace3. Proprietary and Confidential.

Admin Enhancements You Asked For View Manager Admin UI ported to Adobe Flex Dashboard View Reporting DB Delegated Admin Security Server Setup Improvement

Desktop Administrator User location and Filters Individual VM view - Correlation with vCenterpage 99 | 2010 Trace3. Proprietary and Confidential.

Integrated Dashboard UI

page 100 |


System Auditing and Monitoring

page 101 |


Federated Pool Management Feature introduction Total number of VMs can be architected by View Manager(s) up to 10,000 (per vCenter) Number of brokers: 5 + 2 (redundancy) DMZ Security server: 1 + 1 (redundancy) External/internal users: 20/80 Concurrent online users: 100% Maximum sustained rate of logons: 5 per second across all brokers

Concepts Behind the Feature Management was the big cost when scaling. Federated Pool Management: Each VM managed by one broker only

Adding brokers does not increase management load.page 102 | 2010 Trace3. Proprietary and Confidential.

View Connection Servers (View Manager)

page 103 |


View Security Server Pairing Enable security servers to be automatically paired with their broker at installation time Enable configuration changes RDP made on the Broker to be propagated to the security serverCentralized Virtual Desktops

Microsoft Active Directory

vCenter View Connection Server

SSO

Teaming Security Server ( s )

View Client

page 104 |


Transfer Server and Transfer Server Repository The Transfer Server is a new View server role Required for checking out desktop if you plan to use Local Mode Installed in a VM with access to the datastores containing the desktop VMs Windows 2003 and 2008 32bit / 64bit Requires LSI parallel disk controller Stateless without UI using JMS Managed by View Manager Contains an Apache installation as client facing interface to read and write desktop data Multiple Transfer Servers can be used for scalability

Transfer Server Repository is a customer supplied UNC file share View Composer based pools requires Transfer Server for local modepage 105 | 2010 Trace3. Proprietary and Confidential.

Faster checkout

ThinApp Management Feature introduction Associate ThinApp assignment and delivery at desktop pool level Once entitled, it supports visibility into ThinApp status on desktops Event auditing

Benefits Integrates View and ThinApp functionality Providing ease of management and delivery capabilities Dashboard overview of current ThinApps in a client environmentpage 106 | 2010 Trace3. Proprietary and Confidential.

View Composer

page 107 |


View Composer Updated Features Support for SysPrep Refresh, Recompose and Rebalance for Non-Persistent Pools Tiered Storage Support Persistent Disk Management Detach/Reattach/Archive

page 108 |


Sysprep Support Feature introduction Support Sysprep and Quickprep for linked clone guest VMs

Why Sysprep Supported by Microsoft as the only customization method Generates each VM with unique SID Some software (NAC, AV etc) might requires unique SID for licensing control

Restrictions Once a pool is configured either Sysprep or Quickprep, cannot be changed Sysprep is only supported if the pool is using vSphere mode (homogeneous clusters of 4.0 or higher ESX servers)page 109 | 2010 Trace3. Proprietary and Confidential.

Recompose will generate a new SID for a VM created

Sysprep Support

page 110 |


Sysprep Support : Installation and Configuration Installation Install Sysprep on the vCenter server Sysprep functionality is built into the Vista or Win7 OS

View Manager Pool must be configured to use vSphere mode

Linked clone master image View agent with View Composer option must be installed The master image does not need to be joined to the domain For Win7, Volume Licensing must be configured (Microsoft Key Management System server or Multiple Activation Key)page 111 | 2010 Trace3. Proprietary and Confidential.

View Composer Storage Savings View Composer / View achieves storage cost reduction through: Allow storage over-commit Uses delta disks for OS disks and thin provisions user data disks Control the growth of storage via rebalance

page 112 |


Persistent Disk Management Persistent disks are now the first class objects by View Manager Why? When VMs in dedicated pools were deleted, user data could be lost if the persistent disks were deleted Administrators have options to save user data disks (persistent disks) and manage them

Restrictions Recreate desktop can only be done with vSphere pools Attach Persistent Disk can only be attached to VMs in vSphere pools By default disks are archived at the root level of the same datastore they are in. This can be changed if only a single disk is archived.page 113 | 2010 Trace3. Proprietary and Confidential.

Persistent Disk Management: Screen shots Attached Disks

page 114 |



User Experience

View Client PCoIP Protocol Offline

Reduce IT Costs


Management



Platform


page 115 |


O p ti i d C l u d I fra stru ctu re P l tfo rm m ze o n a Scalability: Built for the largest desktop environments 1000s of VMs/pod Faster and more efficient vMotion leading to decreased migration time for VMs Shrink and grow desktops based on demand and priority Dynamic Resource Allocation

High Performance Optimized for desktop workloads Performance acceleration due to lower memory swapping

Best Density Increased desktop VM density 16-20 VMs/core

High Availability and Business Continuitypage 116 | 2010 Trace3. Proprietary and Confidential.

Simplified AV with vShield Endpoint Improve performance and effectiveness of existing endpoint security solutions Offload AV activity to Security VM (SVM) Eliminate desktop agents and AV storms Enable comprehensive desktop VM protection Centrally manage AV service across VMs with detailed logging of AVSVMAV

VMPersona APP OSKernel BIOS



Hardened

OS

Introspectio n

VMware vSphere

page 117 |


Tiered Storage What is tiered storage? Place replicas on a single datastore separate from linked clones The replicas can be shared by all linked clones

Why SSD for Replicas? Use high-performance solid state disks (SSDs) to create replicas Dynamically improve performance of linked clones

Notes vSphere mode only (All ESX servers are 4.0 or higher) Only a single datastore can be selected for replicas The datastore for replicas must be connected to all ESX hosts in the cluster Use with caution as the replica datastore creates a single point of failurepage 118 | 2010 Trace3. Proprietary and Confidential.

Tiered Storage

page 119 |


Tiered Storage: Administration The datastore used for replicas can be changed - it will only affect Newly created VM Recomposed VM Rebalanced VM

If a separate datastore for replicas is de-selected and cannot be found The new VMs, recomposed, or rebalanced VMs will use the OS datastores for replicasReplicas Linked Clone DT OS/Delta 1 Replicas Base Images

View 4.5: Delta Training Revision 1.0

page 120 |


View Composer Feature: Disposable Disk What is disposable disk? Windows and Windows applications writes temporary/paging files to disk These updates are usually deleted after use and space be reused by guest OS

Why? Before, the volume couldnt be reclaimed by VMFS When refreshed on linked clone, it caused persistent data in C:\ drive lost

Benefits: Provides a zero impact to user and lightweight method to reclaim disk space of OS paging file and temporary files.

page 121 |


View Composer: Disposable Disk Disposable Disk Redirect paging and system temp files to a temporary disk removed upon VM powered off Floating View Composer Desktop Dedicated View Composer Desktop

page 122 |


Extensibility with Location Based Printing Leverage ThinPrint AutoConnect dll communicates over a virtual port with a ThinPrint .print Client .print client query locally connected printers and network printers Filter out certain network printers based on the location of the Client Host Integrated via GPO editor in .ADM directory

page 123 |


Extensibility with PowerShell Provides a series of PowerShell cmdlets to administer View from the command-line (PowerCLI) ## Linked Clone operations accept individual machine ids. ## The below commands can be used to cover all the VMs Allows management of: in a pool. VI server entries View Licenses Get-DesktopVM -pool_id | SendLinkedCloneRecompose -schedule (Get-Date) Global Config -parentVMPath Remote and Local Desktop Sessions Desktops/Pools VMs and Physical Machines (running the Agent) EntitlementsGet-DesktopVM -pool_id | SendLinkedCloneRebalance -schedule (Get-Date) Get-DesktopVM -pool_id | Send-LinkedCloneRefresh -schedule (Get-Date)

Why? Allows for automation and scripting Provides extensibility to Administration tasks Seamless integration from View to vCenterpage 124 | 2010 Trace3. Proprietary and Confidential.

Extensibility in Core Broker View Framework SDK A backend consolidation that extends vdmadmin PowerShell cmdlets defined on the .NET bridge SCOM support

page 125 |


Extensibility with Kiosk Mode Locked down View Access Client device ID based provisioning and auto-logon Automatic generation of ClientID based user account in AD

Kiosk-ready View Client Suppression of GUI features Error reporting for script integration

Automated USB redirection Client info support for in-guest printer mapping ThinPrint GPO enabled for location based printing Use cases: airport check-in, library, amusement park event kiosk, registration desk, ticketing

page 126 |


Extensibility with GPO Templates Control View components behaviors domain-wide Configure location-based printing Creating an OU for View desktops vdm_agent.adm (allow protocol access, SSO, run commands etc) vdm_client.adm (pass endpoint client information to agent etc) vdm_server.adm (performance and log configuration etc) vdm_common.adm (common configuration) pcoip.adm (limit peak bandwidth)page 127 | 2010 Trace3. Proprietary and Confidential.

page 128 |


Thin Client Overview

WYSE

page 129 |


Flexible thin clients

Desktop Thin Clients Lowest power (

View Roadshow

Documents

Transcript of View Roadshow