At TechEd Europe 2014 in Barcelona, the independent research group Project Virtual Reality Check (Project VRC) announced the release of a publicly available white paper with performance results and best practices for Microsoft App-V 4.6 and 5.0. The white paper —based on lab environment testing— contains detailed App-V performance best practices that system administrators and IT managers in mid- to large-sized enterprises can apply to any new or existing App-V infrastructure.
Microsoft App-V is the leader in the application virtualization market. Previous surveys by Project VRC showed that almost 35% of the surveyed organizations are using Microsoft App-V in their virtualized desktop environment. This new independent white paper focuses on the performance impact and best practices of Microsoft App-V 4.6 and 5.0 in virtualized desktop environments. Enterprises may use these best practices to better plan their infrastructures that include App-V in order to deliver excellent performance to end users, while reaping the benefits of moving applications to centrally managed services.
When comparing the capacity impact of App-V to locally installed applications in a virtualized desktop environment, App-V 5.0 will cause a capacity impact of 13%. In other words, 13% fewer users can be supported on any given hosted desktop environment. In the same scenario, App-V 4.6 will only have an impact of 8% in comparison to locally installed apps. Research by Project VRC shows that this difference of 5% is caused by the fact that App-V 5.0 has more and tighter integration options with the local operating system.
Streaming applications also has a negative impact of 15% (through HTTP) and 6% (through SMB) compared to locally cached applications. This offset is also noticeable in publishing times. A small impact (4%-7%) on capacity was measured when applying Registry Staging; both in cached as well as in a streaming scenarios.
Overall, user-based publishing is 20-25% faster than global publishing. App-V 5.0 Hotfix 4 results in a big improvement in publishing times (around 38%). There’s no impact on capacity. Project VRC also shows that enabling Shared Content Store will minimize the write I/O on the underlying storage system drastically, resulting in 16% more capacity compared to a traditional streaming scenario.
Ryan Bijkerk, Development Manager at Login VSI, says “Another key finding we discovered is that .NET optimization needs to be enabled when using Microsoft .NET Framework 4.0 with App-V 5.0 at all times. When App-V administrators are not aware of this, they can encounter many performance problems.”
Ment van der Plas, Microsoft MVP for App-V, says: “Besides testing our own scenarios, we decided to validate already known best practices and discovered some interesting results. Commonly known and heavily debated performance impacting sequencing guidelines proof to have little to no impact in our test results.”
All Project VRC tests are performed with Login VSI software. This vendor-agnostic tool simulates realistic user workloads to objectively test the performance of virtualized desktop environments. The full test methodology used is described in the white paper.
About Project VRC
Project Virtual Reality Check (Project VRC) was started in 2009 by two companies—PQR and Login Consultants—with deep expertise in VDI and SBC. Project VRC’s mission is to deliver insightful and useful independent research on desktop virtualization. Several white papers were published about the performance of different hypervisors, application virtualization solutions, Windows operating systems, antivirus solutions and Microsoft Office versions. Visit www.projectvrc.com for more information.