When doing different performance researches it is important to have a clear understanding of the testing methodology used. With all the information available you must be able to reproduce the exact same results. In case of a performance research, we use various tools to measure the impact from the performance perspective. The reproducibility is very important to publish reliable and consistent results. The following section will describe some details about the tooling, data sources, and automation.
Software and tools
The primary tool to generate a reliable load on the environment is done by the product Login VSI. Login VSI is an industry standard in simulating users on virtual desktop environments. Each performance research will be tested using Login VSI unless there is a reason to use another solution.
For all organizations using centralized virtual desktop environments to provide applications to their end-users, we offer the Login VSI Enterprise Edition, the complete solution to protect VDI performance and availability.
Login VSI offers a unique combination of synthetic load-testing and pro-active monitoring capabilities, allowing enterprises to design, build and maintain VDI environments (both infrastructure and applications) that can provide, and safeguard, an optimal End-User Experience.
The Login VSI load-testing solution generates a large number of synthetic users to test and protect the performance and scalability of your new and existing VDI, SBC and DaaS deployments.The Login VSI is the complete solution to optimize and protect the performance of business-critical applications running in SBC and VDI infrastructures such as VMware Horizon View, Citrix XenDesktop, Citrix XenApp, and Microsoft Remote Services (RDS), previously Terminal Services.
The default LoginVSI settings do not always suit our needs. Consequently, we’ve made some changes to the test runs. The default PDF printer is DoroPDF which is not always reliable under maximum load which results in stuck sessions. As this effect the results we decided to disable PDF printing by default to ensure a smooth execution of the workload.
Data source and testing methodology
The default LoginVSI settings do not always suit our needs. Consequently, we’ve made some changes to the test runs. The default PDF printer is DoroPDF which is not always reliable under maximum load which results in stuck sessions. As this effect the results we decided to disable PDF printing by default to ensure a smooth execution of the workload.
For each scenario, the test cycle is repeated 10 times. All performance results published on the ICT-R platform are based on multiple test cycles and reported in the average. This way we can ensure the findings of the results are consistent and correct.
We hope this provides enough insights into the testing methodology used in the ICT-R lab environment. If you have any comments or question please leave them below.
ict-r.com - information technology
