Scope
One of the principal advantages of deploying multiple users on a single host–be it a PC, physical server or virtual machine–is that you can readily scale the size of your deployment as your users’ needs grow. Rather than being forced to purchase expensive new computing resources for each new user, you can simply expand the capacity of your host system. vSpace Server 6 now allows you to extend the number of users per server up to 100 users depending on your user workload, host system performance, and client devices being used.The number of users, intended application suite and overall performance expectation determine how powerful a host system must be in order to deliver the desired end-user experience. This document provides high-level guidelines for determining the system requirements for various numbers of L-series and M-series users with various computing workloads. This document can be used as a starting point for sizing your deployment – but your own in-house testing should be used for the final determination of your host systems’ configurations.
Understanding Use Cases for General Usage and Video Playback Workloads
Prior to deploying vSpace, you should develop use cases for the users you expect to connect to a host. An important consideration in developing use cases involves determining the number and types of applications users will need to operate. These requirements help you identify and measure the users’ standard workloads. For example, you should measure the CPU, memory and storage utilization for a typical user workload in your environment. This workload data and the total expected number of users will help you determine the system requirements for your host system.
An important concept to keep in mind is the difference in resource usage seen when comparing ordinary office application usage and multimedia streaming. As demonstrated in the chart below, our Core i7 test system was able to support roughly 60 users whose workload was comprised entirely of common office application usage. Once the use-case switched to video streaming however, the number of users that could be adequately supported by the same hardware specs dropped to 20. This is a good example of how demanding video streaming and other similar tasks (such as browsing flash intensive websites) can be on a system. With regards to video streaming, another defining variable is the size of the video being streamed. For example, 720p video contains approximately three times (3x) the amount of pixel data as 480p video. Understandably that translates to a notable difference in resource requirements between the two formats.
For the purpose of our testing, workloads were divided into three typical usage scenarios each of which involved sequentially cycling through and completing tasks within the listed applications. These workload profiles have been detailed below. We've also included test data for two different levels of video fidelity which represent "extreme" usage scenarios wherein all users present were streaming videos. Naturally, you are not required to build your host specifications around extensive high definition video streaming unless that is a core requirement for your deployment. Read more tech details here....
