Dell Wyse Datacenter For VMware Horizon DaaS And Silver Peak

1y ago
18 Views
2 Downloads
701.58 KB
16 Pages
Last View : 29d ago
Last Download : 6m ago
Upload by : Aliana Wahl
Transcription

Dell Wyse Datacenter for VMware HorizonDaaS and Silver PeakWAN optimization testing for a DaaS environmentDell Wyse Solutions EngineeringJune 2014A Dell Technical White Paper

Revisions2DateDescriptionJune 2014Initial release v.1.0Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

THIS WHITE PAPER IS FOR INFORMATIONAL PURPOSES ONLY, AND MAY CONTAIN TYPOGRAPHICAL ERRORS ANDTECHNICAL INACCURACIES. THE CONTENT IS PROVIDED AS IS, WITHOUT EXPRESS OR IMPLIED WARRANTIES OFANY KIND. 2014 Dell Inc. All rights reserved. Reproduction of this material in any manner whatsoever without the expresswritten permission of Dell Inc. is strictly forbidden. For more information, contact Dell.PRODUCT WARRANTIES APPLICABLE TO THE DELL PRODUCTS DESCRIBED IN THIS DOCUMENT MAY BE FOUNDAT: ommercial-and-public-sector Performance of networkreference architectures discussed in this document may vary with differing deployment conditions, network loads, andthe like. Third party products may be included in reference architectures for the convenience of the reader. Inclusionof such third party products does not necessarily constitute Dell’s recommendation of those products. Please consultyour Dell representative for additional information.Trademarks used in this text:Dell , the Dell logo, Dell Boomi , Dell Precision ,OptiPlex , Latitude , PowerEdge , PowerVault ,PowerConnect , OpenManage , EqualLogic , Compellent , KACE , FlexAddress , Force10 and Vostro are trademarks of Dell Inc. Other Dell trademarks may be used in this document. Cisco Nexus , Cisco MDS , Cisco NX 0S , and other Cisco Catalyst are registered trademarks of Cisco System Inc. EMC VNX , and EMC Unisphere are registered trademarks of EMC Corporation. Intel , Pentium , Xeon , Core and Celeron are registered trademarks of Intel Corporation in the U.S. and other countries. AMD is a registered trademark and AMD Opteron , AMD Phenom and AMD Sempron are trademarks of Advanced Micro Devices, Inc. Microsoft , Windows , Windows Server , Internet Explorer , MS-DOS , Windows Vista and Active Directory are either trademarks or registered trademarks of Microsoft Corporation in the United States and/or other countries. Red Hat and Red Hat Enterprise Linux are registered trademarks of Red Hat, Inc. in the United States and/or other countries. Novell and SUSE are registered trademarks of Novell Inc. in the United States and other countries. Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Citrix , Xen , XenServer and XenMotion are either registered trademarks or trademarks of Citrix Systems, Inc. in the United States and/or other countries. VMware , Virtual SMP , vMotion , vCenter and vSphere are registered trademarks or trademarks of VMware, Inc. in the United States or other countries. IBM is a registered trademark of International Business Machines Corporation. Broadcom and NetXtreme are registered trademarks of Broadcom Corporation. QLogic is a registered trademark of QLogicCorporation. Other trademarks and trade names may be used in this document to refer to either the entities claimingthe marks and/or names or their products and are the property of their respective owners. Dell disclaims proprietaryinterest in the marks and names of others.3Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

Table of contentsRevisions . 2Executive Summary . 5VMware Horizon DaaS . 512Configuration of the test environment . 61.1Hardware component details . 61.2Software component details . 71.3PCoIP configuration details. 71.4Networking configuration diagram . 8Test results and analysis . 92.1Test case 1 . 92.1.1 Application bandwidth reduction . 102.1.2 Peak appliance throughput . 112.2Test case 2 . 122.2.1 Application bandwidth reduction . 122.2.2 Peak appliance throughput . 1332.3Test case 3 . 152.4Results summary . 15Conclusion . 16About the authors . 164Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

Executive SummaryThis document contains the architecture and results for testing conducted to investigate the impact ofWindows 7 on VMware Horizon DaaS 6.0 using Ivy Bridge processors, and determine the functionality ofthe Silver Peak wide area network (WAN) optimization appliances in an emulated WAN environment.VMware Horizon DaaSVMware Horizon DaaS enables companies to deliver virtual workspaces to their end users – including fullWindows Client desktops, shared desktops and applications – as a monthly subscription serviceVMware enables a risk-free incremental evolution of the next-generation workspace. It provides acomplete virtual workspace from the cloud, delivering desktops and applications as an easily-managed,integrated cloud service. Enterprises can rapidly provision desktops and applications to users on anydevice, anywhere. And virtual desktops are transformed from capital expenditures inherent in onsiteenterprise desktop virtualization and physical PC refreshes, into a predictable operating expenditure.5Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

1Configuration of the test environment1.1Hardware component detailsHardware Components:VMware ComputeHost1 x Dell PowerEdge R720 Servers: VMware ESXi 5.1 Update 2 2 x E5-2690 v2 @ 3.0 GHz 10c IvyBridge Processors 256GB Memory @ 1600MT/s 12 x 300GB 15K SAS internal diskdrives (Raid 10) Broadcom BCM5720 1GbE QP NIC iSCSI: Broadcom 57810 PERC H710P RAID Controller BIOS version: 2.0.18 (or higher)VMwareManagement Host1 x Dell PowerEdge R720 Servers: Test Launcher Server(VDI WorkloadGenerator)2x Dell PowerEdge R710 servers: Network 6VMware ESXi 5.1 Update 22 x E5-2690 v2 @ 3.0 GHz 10c IvyBridge Processors256GB Memory @ 1600MT/s12 x 300GB 15K SAS internal diskdrives (Raid 10)Broadcom BCM5720 1GbE QP NICiSCSI: Broadcom 57810PERC H710P RAID ControllerBIOS version: 2.0.18 (or higher)VMware ESXi 5.5Intel-(R)-Xeon- (R) CPU X5670 @2.9 GHz96 GB RAM @ 1333 MHZ10 x 146GB 15K SAS internal diskdrivesBroadcom 5709 1GbE NICLogin VSI version 4.11 x Dell PowerConnect 6248 1GbEthernet Switch2 x Dell Force10 S4810 switches3 x WANemDell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0DescriptionThis system hosts the Windows 7VDI desktops.For the ESXi environment, 12 x 146GB drives were configured in aRAID 10 array.Each VM hosted the followingworkloads on Windows server2008 R2 SP1 and Server 2012 R2: VMware vCenterFile ServerDomain ControllersStratusphere Hub (5.6.2)VMware Horizon DaaSappliancesVMware ESXi 5.5 installed on theTest Mgmt R710 server.Login VSI launchers created onthis server to initiate VMware Viewsessions from each VM tosimulate VDI workload.The VMs were loaded withWindows 8.1 Enterprise 64-bitoperating systems.Deployment Stack and Test Stacksconfigured on separate switchesseparated by physical switchesand virtual WANem appliances.

1.2PerformanceMonitoring VMware Virtual Center 5.1Silver Peak GMSPerformance data captured fromVSphere client and GMS webinterface using scheduled reportsWAN Optimization Silver Peak GX-V v6.2.4Silver Peak VX-7000v6.2.4.0 49732The VX-7000 appliances will beused to optimize WAN trafficSoftware component detailsSoftware ComponentsDescription/VersionVMware vCenterVMware Horizon DaaSHypervisorWindows 7 Enterprise (32-bit)Windows 8.1 Enterprise (64-bit)Windows Server 2012 R21.3Version 5.1Version 6.0VMware ESXi 5.1 Update 2Virtual DesktopsLogin VSI launchersInfrastructure VMsPCoIP configuration detailsThe gold image used for testing was configured using VMware’s PCoIP best practices. These settings areshown on page 20 in the “VMware Horizon View Best Practices Performance Study” white paper foundhere. The settings are intended for VMware Horizon View environments, but also apply to VMware HorizonDaaS desktop images using PCoIP. The following PCoIP settings were applied to the gold image in thetest environment, but it should be noted that the optimal settings for any other deployment may varygreatly depending on the operating environment, WAN bandwidth and quality of the link.7SettingDefaultConfiguredDescriptionBuild to losslessOnOffEnabled ability to enable or disable build tolosslessMaximum sessionbandwidth90,000 Kbps90,000 KbpsSpecifies the maximum bandwidth in aPCoIP sessionPCoIP session bandwidthfloorNo minimum0Specified a lower limit, in kilobits persecond, for the bandwidth that is reservedby the PCoIP session.Session Audio BW limit500 Kbps100 KbpsReduces bandwidth usage of audio withusable qualityMinimum Image Quality5050Maximum Initial ImageQuality9070Maximum Frame rate3015Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0In WAN conditions, this will be helpful forvideo playback and fast graphics operations

1.4Networking configuration diagramThe following network diagram shows the logical layout of the test environment. There are geographicallyseparated service provider, data center, and client sites. Between each of the sites, are Silver Peak andWAN emulation (WANem) appliances to simulate a WAN environment.8Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

2Test results and analysisThe primary focus of the tests was to determine the effect of using Silver Peak WAN optimizationappliances with Windows 7 using VMware Horizon DaaS 6.0 on Ivy Bridge processors. Aspects that wereinvestigated were bandwidth reduction, error correction, latency, and throughput. In each test, 110Enhanced users were used as this desktop density was well within the capabilities of the PowerEdge R720compute host and SolidFire SF-3010 storage array at the VMware DaaS Service Provider site. Note that 110Login VSI Enhanced user workloads were not chosen to test scale of server or WAN but rather to generatean application WAN traffic payload across the back haul to the corporate data center from the desktopinfrastructure in the cloud.Primary testing objectives: Validate that the Silver Peak WAN optimization (WANOP) appliances function as expectedDetermine what affect the WANOP devices have on user experience and performanceDetermine what affect the WANOP devices have on application trafficThe following table lists the tests that were performed and the parameters that were variable betweentests. The Service Provider (SP), Customer Datacenter (DC), and Client Site (CS) latencies and loss weremodified as follows:Test #SP to DC LatencyCS to SP Latency140 ms @ 1% loss38 ms280 ms @ 1% loss38 ms340 ms @ 1% loss120 ms @ 2 % lossNote: Test 3 is quantified by LoginVSI client-side test tools (or other client-side test tools)In all tests, the Silver Peak maximum bandwidth was set to 155,000 Kbps. Also, forward error correction(FEC) setting was set to “enable” with the default ratio of 1:10. The FEC setting is described in the followingscreenshot:Each test was conducted twice. Once with the Silver Peak appliance optimizations disabled and once withthe optimizations enabled to determine the effect of the optimizations on network traffic and desktopexperience.2.1Test case 1The following validation was done for 110 Enhanced users using the standard Performance Analysis andCharacterization (PAAC) testing methodology called out in the Reference Architecture document.Test # 19SP to DC Latency - 40 ms @ 1% lossDell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0CS to SP Latency - 38 ms

2.1.1Application bandwidth reductionAs seen in the following diagrams, LAN (light blue) and WAN (dark blue) bandwidth are the same in thediagram on left. This indicates that all data transferred on the LAN also had to traverse the WAN to accessthe file or web services.In the diagram on the right, WAN traffic is greatly reduced while LAN traffic remains very high. Thisconfirms that the Silver Peak WAN appliances served the CIFS and HTTP resources out of cache andreduced the WAN traffic by 98%.PCoIP traffic remains unchanged as a major portion of the packets are encrypted and look unique fromthe perspective of the Silver Peak WAN optimization appliances and thus are not served from cache.The diagrams below depict the application reduction by time for the duration of the test. As seen in thediagram on the left, there is no reduction of data and the LAN and WAN traffic are nearly identical. In thediagram on the right, we can see that the data that traversed the WAN was drastically reduced while theLAN traffic remained very high.10Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

2.1.2Peak appliance throughputIn the diagrams below, the peak throughput can be seen on both the LAN and WAN for all appliances. Inthe diagram on the left, peak throughput for the WAN is 140 Mbps while the LAN is approximately 153. Thepeak throughput is very close in this scenario as the data is not cached on the Silver Peak appliances.In the diagram on the right, the peak WAN throughput is less than 10 Mbps while the LAN is actuallyslightly faster than the gigabit Ethernet that was used for the test. This is most likely due to the fact that theSilver Peak VX-7000 was caching the CIFS and HTTP traffic and was connected to the same VMwarevSwitch as the tenant desktops.The following diagrams, the throughput over the WAN and LAN are shown for the duration of the test.Figure 111Silver Peak WAN optimization disabledDell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

Figure 22.2Silver Peak WAN optimization enabledTest case 2The following validation was done for 110 Enhanced users using the standard PAAC testing methodologycalled out in the Reference Architecture document.Test # 22.2.1SP to DC Latency - 80 ms @ 1% lossCS to SP Latency - 38 msApplication bandwidth reductionAs seen in the following diagrams, LAN (light blue) and WAN (dark blue) bandwidth are the same in thediagram on left. This indicates that all data transferred on the LAN also had to traverse the WAN to accessthe file or web services.In the diagram on the right, WAN traffic is greatly reduced while LAN traffic remains very high. Thisconfirms that the Silver Peak WAN appliances served the CIFS and HTTP resources out of cache andreduced the WAN traffic by 80%.PCoIP traffic remains unchanged as a major portion of the packets are encrypted and look unique fromthe perspective of the Silver Peak WAN optimization appliances and thus are not served from cache.12Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

The diagrams below depict the application reduction by time for the duration of the test. As seen in thediagram on the left, there is no reduction of data and the LAN and WAN traffic are nearly identical. In thediagram on the right, we can see that the data that traversed the WAN was drastically reduced while theLAN traffic remained very high.2.2.2Peak appliance throughputIn the diagrams below, the peak throughput can be seen on both the LAN and WAN for all appliances. Inthe diagram on the left, peak throughput for the WAN is 93 Mbps while the LAN is approximately 90. Thepeak throughput is very close in this scenario as the data is not cached on the Silver Peak appliances.In the diagram on the right, the peak WAN throughput is less than 10 Mbps while the LAN is near speedsnative to the gigabit Ethernet that was used for the test. This is most likely due to the fact that the SilverPeak VX-7000 was caching the CIFS and HTTP traffic and was connected to the same VMware vSwitch asthe tenant desktops.13Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

The following diagrams, the throughput over the WAN and LAN are shown for the duration of the test.Figure 314Silver Peak WAN optimization disabledDell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

Figure 42.3Silver Peak WAN optimization enabledTest case 3The following validation was done for 110 Enhanced users using the standard PAAC testing methodologycalled out in the Reference Architecture document.Test # 3SP to DC Latency - 40 ms @ 1% lossCS to SP Latency - 120 ms @ 2% lossThis test was very similar to test case 1, but the latency and loss setting between the service provider andcustomer sites were modified to determine the effect that the Silver Peak appliance had on PCoIPpresentation traffic.The following table details the findings that have been averaged over all of the desktops:Desktop Login Delays2.415SP DisabledSP Enabled% Improvement (round)117.3851.54 128%Results summaryTestCaseCS to SPLatencyCS to SPLossSP to DCLatencySP toDC LossSilver Peak DataReductionMax. Throughput(Not Optimized)Max. Throughput(Optimized)138040198%153 Mbps1120 Mbps238080180%93 Mbps926 Mbps3120240178%145 Mbps899 MbpsDell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

3ConclusionThe Silver Peak VX-7000 WANOP appliances greatly improves the quality of line and the resulting desktopexperience for customers who wish to utilize the Dell Wyse Datacenter for VMware Horizon DaaS solutionwithin a service provider scenario. The appliances were shown to reduce WAN traffic by up to 98% anddecrease user login delay by up to 56%.About the authorsNicholas Busick is a Senior Solutions Engineer with Dell Wyse Solutions Engineering building, testing,validating and optimizing enterprise VDI stacks.Gus Chavira is the Senior Principal Engineering Architect for VMware Horizon based solutions at Dell. Gushas extensive experience and expertise on the VMware solutions software stacks as well as in Enterprisevirtualization, storage and enterprise data center design. Gus has worked in capacities of Sys Admin, DBA,Network and Storage Admin, Virtualization Practice Architect, Enterprise and Solutions Architect. Inaddition, Gus carries a B.S. in Computer ScienceDavid Hulama is the Senior Technical Marketing Advisor for VMware Horizon View solutions at Dell. Davidhas a broad technical background in a variety of technical areas and expertise in enterprise-classvirtualization solutions.16Dell Wyse Datacenter for VMware Horizon DaaS and Silver Peak v.1.0

The gold image used for testing was configured using VMware's PCoIP best practices. These settings are shown on page 20 in the "VMware Horizon View Best Practices Performance Study" white paper found here. The settings are intended for VMware Horizon View environments, but also apply to VMware Horizon DaaS desktop images using PCoIP.

Related Documents:

5- Wyse ThinOS 8.6 is ready to connect to Citrix , Microsoft and VMware environments. Wyse ThinOS 8.6 with PCoIP is required for accessing Amazon Workspaces and VMware environments using PCoIP. Wyse ThinOS 9.0 provides exclusive access to Citrix Workspace including Citrix Virtual Apps and Desktops Wyse ThinOS 9.0 requires Wyse Management .

Wyse Datacenter for Citrix XenDesktop Reference Architecture v.7 2 Solution Architecture Overview 2.1 Introduction The Dell Wyse Datacenter Solution leverages a core set of hardware and software components consisting of five primary layers: Networking Layer Compute Server Layer Management Server Layer Storage Layer

Dell Wyse TCX/VDA PC Extender Xenith 2 Xenith Pro R10L T10 C10LE Z50D/S D50D T50 C50LE X50m Citrix HDX Citrix PS Dell Wyse TCX/VDA Dell Wyse WSM Z90DW/7 Z90SW/7 Z90DE7 D90DW/7 C90LEW/7 X90mw/7 X90cw/7 View RDC7** Dell Wyse

Meet the Wyse Family Wyse ThinOS Enjoy peace of mind with Dell's exclusive Wyse ThinOS, the most secure thin client operating system3. With a closed architecture, system hardening to limit attack surfaces, an unpublished API, encrypted data and files exclusively packaged by Dell, Wyse ThinOS is virus and malware resistant.

Dell Wyse Cloud Connect Designed to promote bring-your-own-device (BYOD) environments, Dell Wyse Cloud Connect allows you to securely access and share work and personal files, presentations, applications and other content from your business or your home. Managed through Dell Wyse Cloud Client Manager software-as-a-service (SaaS), IT

Wyse thin clients with Windows 10 IoT Enterprise optimized by Dell provide enhanced user experience, increased uptime and stronger security. Experience confident productivity with Wyse Overlay Optimizer. This unique solution provides write protection and extends the uptime of Wyse thin clients. Wyse Overlay Optimizer is a patent pending

The Wyse 5010 (ThinOS) is Citrix HDX, Microsoft RemoteFX, and VMware Horizon View certified. It also supports legacy peripherals via an optional USB adapter. For more information, please visit: Link Wyse 5010 Thin Client (Windows Embedded Standard 8) A strong, reliable thin client, the Dell Wyse 5010 thin client runs Windows

2.7 VMware vCenter Support Assistant 22 2.8 VMware Continuent 23 2.9 VMware Hyper-Converged Infrastructure Kits 23 2.10 VMware Site Recovery Manager 23 2.11 VMware NSX 24 2.12 VMware NSX Advanced Load Balancer 28 2.13 VMware SD-WAN by VeloCloud 29 2.14 VMware Edge Network Intelligence 30 2.15 VMware NSX Firewall 30