One office with one type of desktop… Easy. Hundreds of offices with any type and age of desktops… Difficult but not impossible.
Most organizations find themselves in the difficult camp. A user’s desktop can be completely different (in terms of hardware, resources, applications and configuration) than the person sitting next to them doing a similar job. As the environment includes users from different departments, in different offices, with different requirements it becomes clear that the understanding of the user topology for an organization is critical before one can create a desktop virtualization solution.
In previous blogs, I’ve discussed how understanding the underlying standards, applications and storms plays an important role in creating a successful virtual desktop design. The fourth requirement is to understand the organization’s user topology. More specifically, one must get a grasp of the endpoints and user locations.
First, the endpoints. Most organizations follow a 3-5 year desktop refresh cycle. At a minimum, there will be 5 different hardware configurations for each of the 5 years (in actuality, there will likely be many, many, many more configurations). Also, the desktops that are less than 2-3 years old have hardware configurations that can easily support Windows 7 and the latest applications. These newer desktops have more virtual desktop options than an endpoint that is 5+ years old. Newer desktops have the processing power to support the Local Streamed Desktop FlexCast model instead of the hosted VM-Based VDI desktop model.
With Local Streamed Desktop, the desktop is still virtualized and centrally managed, the desktop still receives the virtualized applications, and the users still have their personalized settings applied. The difference is that instead of using resources on a physical server in the data center, the local desktop resources are used. Because local desktop resources are consumed, fewer data center servers are required to support the same number of users
This is but one example of how understanding the endpoints helps determine the type of virtual desktop a user requires. However, just knowing the endpoints is only one aspect of the user topology. The second aspect, user’s location, also plays an important role in selecting the most appropriate virtual desktop.
Certain desktops require a high-speed connection to the infrastructure while other options can allow slower networks with higher latency. By assessing the user locations and the connections to the data center, the proper solution can be put into place to support the virtual desktop FlexCast model.
- Hosted shared desktop: Can be used on networks with low speeds and high latency
- Hosted VM-based VDI desktop: Can be used on networks with low speeds and high latency
- Hosted blade PCs: Can be used on networks with low speeds and high latency
- Streamed local desktop: Requires a fast, low latency network to the physical desktop for optimal performance
- Virtual Apps to Installed Desktops: Can be used on networks with low speed and high latency. If application streaming is used (as compared to hosted applications), slower networks will delay application startup time, but users have the ability to work disconnected.
- Local VM-based desktop (not yet available): Can be used on networks with low speed and high latency, although the slower the network the longer it will take to sync the image to the endpoint. Images can be tens of GBs in size. But once delivered to the end point, all communication remains local to the desktop.
When deciding on the appropriate virtual desktop type, the endpoint and the user’s location matter. Without taking both into account, a user might end up with a fast virtual desktop that takes 5 minutes to start Microsoft Word. Gather all the information before deciding on your virtual desktop type.
Lead Architect – Worldwide Consulting Solutions
Follow Me on twitter: @djfeller
Blog for Next-Gen Desktop: Ask The Architect
Questions, then email Ask The Architect
Facebook Fan Page: Ask The Architect