35% improvement of VDI user density with HP ProLiant BL460c Gen8 & Citrix XenDesktop 5.6 & PVS 6.1
Author: Kirk Manzer, Technologist and Architect, Client Virtualization Engineering
An increasingly global and mobile workforce is changing the way IT will support the business. By virtualizing the desktop to be accessed by a variety of end user devices (whether it is a Smartphone, tablet, laptop, or thin client), HP and Citrix make this important move to client virtualization safe and easy by testing and developing the most popular solutions
HP has measured the ProLiant BL460c Gen8 with Citrix XenDesktop using LoginVSI and the results are worthy of the ProLiant reputation.
VSI Max score for medium script users – 214 users
VSI Max score for heavy script users – 174 users
Measuring VDI Performance
HP over time has used numerous tools to measure server performance for VDI. Those included internally developed, leveraging partners’ tools and most recently the newest version of LoginVSI which provides useful data when testing system performance.
LoginVSI overview
LoginVSI is a benchmarking tool designed to measure the performance of centralized desktop environments such as Server Based Computing and Virtual Desktop Infrastructure.
Test Environment
All tests were performed on a single HP ProLiant BL460c Gen8 server, Citrix XenDesktop 5.6, and PVS 6.1 running on Microsoft Windows Server 2008 R2 Hyper-V. PVS was configured with client side write cache. An HP D2700 storage enclosure with 6 Gb SAS connectivity was used to provide 6 300GB 15K SAS drives to the server as direct-attached storage in a RAID 1+0 configuration.
The PVS write-cache files are stored on the direct-attached storage.
The server tested featured 2 Intel XEON E5-2680 processors and 256GB of 1333MHz PC10600 RAM. Faster memory is available and recommended as memory speed is more important than capacity in achieving maximum results. Memory running slower than 1333MHz should not be used.
Virtual machines were optimized following Citrix’s recommended procedures, running Microsoft Windows 7 64-bit. All VMs leveraged Microsoft Windows Server 2008 dynamic memory and all VMs were allocated a minimum of 768 MB of memory and a maximum of 1.5 GB of memory. All VMs used a single vCPU.
HP recommends sizing your VMs to your application and usage patterns. Assigning the correct amount of memory is one of the easiest ways to optimize performance in a VDI environment.
Workloads
LoginVSI offers a number of workloads intended to emulate different user behaviors. To best emulate HP’s traditional test methodology, the medium and heavy user workloads were chosen. These workloads represent productivity and knowledge workers respectively. HP starts launchers at a fixed, 35 second interval. All launchers are virtual machines connected to a domain and configured for NTP services.
Detailed information about LoginVSI is available here. In particular, dissemination of how tests are run as well as how scores are calculated can be accessed at the Admin Guide.
HP ProLiant BL460c Gen8 Results: Medium Workload– VSI Max of 214
The medium workload is frequently used to express user counts that represent general productivity users or “average” end user s. HP has chosen to report this workload as being representative of such a user with the caveats listed in the section entitled Real world estimations and recommended sizings. HP strongly recommends consulting that section prior to planning a VDI proof of concept or production implementation.
Using the ProLiant BL460c Gen8 server configured as listed in this document HP received a VSIMax score of 214 users with a medium workload. Response times for actions measured within the script remained very low as shown in the graph below validating the quality of HP’s end user experience.
Figure 1: Medium script VSI Max score.
Heavy Workload
The heavy workload represents workers who place a great deal of pressure on their systems with multiple applications open and heavy interaction on a frequent basis.
Running the heavy workload with the ProLiant BL460c Gen8 configured as described in this document HP achieved a VSIMax score of 150 users with very fast response times on measured script actions.
Figure 2: Heavy user
Baselines
HP has not previously published results using LoginVSI on ProLiant systems of G7 and earlier generations. As a point of reference and to establish a baseline, HP is making those results available for the medium user script. The tests were conducted using an HP Virtual System CV2 and a ProLiant BL460c G7 with 2 X5775 processors and 192GB of 1333MHz PC10600 memory. Using this method HP achieved a VSI Max score of 143 medium users as shown in the figure below.
Figure 3: Baseline results for G7 servers and a medium script workload
This substantiates that the ProLiant BL460c Gen8 has seen a better than 30% performance increase versus the same server in G7 configuration.
Real world estimations and recommended sizings
Prior to making recommendations it is important to note that the following numbers are just guidelines. While these numbers will be achievable by some customers, others may find numbers that are significantly higher or lower. Adherence to best practices during all aspects of the design and build phases of your VDI implementation as well as choosing optimized methods for deploying and managing VDI will nearly always result in greater user density. Selecting problematic applications, poorly suited user groups and failing to follow recommendations will nearly always have the opposite effect. This is critical because these densities have a potentially large impact on the acquisition and long term costs of running VDI in a given environment.
HP recommends the following planning numbers to estimate real world sizing based on the graphs achieved by running LoginVSI. These results are ranges that estimate from the graphs and data obtained from the LoginVSI tests where the user experience begins to change. It has been HP’s experience over many years of working with customers on VDI implementations that the quality and consistency of the end user experience is a large factor in determining end user acceptance as well as long term success. Table 1 shows the VSIMax as well as the recommended planning numbers which are factored as 60-65% of the VSImax score for light and medium scripts and 50-55% for heavy users.
Table1. Recommended planning numbers by user type.
|
User Type |
VSImax |
Recommended Range |
|
Medium |
214 |
128-139 |
|
Heavy |
174 |
87-95 |
There are many variables that will have an effect on your overall sizing including proper image optimization, memory speed and architecture, concurrency ratio of active users to total users, application delivery methods, operating system choice, protocol and provisioning methodology. All sizing numbers should be used as guidelines only.
Contact your HP representative today to discover how HP client virtualization solutions for Citrix Xen Desktop can unleash your workers’ productivity, streamline management, protect your information and lower operating costs.
For more information on HP client virtualization solutions, visit www.hp.com/go/cv
For more information on HP Gen8 servers, click here
