Virtual Desktops and Flying Cars

In the 1950s, there was a widespread expectation that by now we would all be driving flying cars as a matter of routine. With the advances in aeronautics in the 1940s and 1950s, it would not have been hard to imagine that an economic flying car that the general public would be happy to use would be common within far less than 50 years. The reality is that the flying car failed on both counts; personal flying never became economic, and flying, in any form, is a hobby for only a tiny portion of the population.

Back in 2005, you might have predicted that VDI, as desktop virtualization was then called, would be common today as a way of delivering corporate desktops. That has not happened for a number of interesting reasons, which point toward the real future of desktop virtualization and contrast markedly with the history of the flying car. In 2005, it was amazing that you could centrally virtualize a desktop at all. Virtualization was very new for servers and using a hypervisor on a server to host multiple client operating systems was a revolutionary idea that quickly caught peoples’ imaginations. While there were some good use-cases for those early deployments, it did not make sense to manage the majority of users this way. The problem was that simply moving the PC image into the data center did not save money and in many cases increased capital costs without decreasing operational costs. What was needed was a better way of managing PCs, rather than just a way of running them on hypervisors in the data center.

Essentially, the problems in managing PCs stem from the "personal" nature of the PC. While we may deploy a gold image to new machines during hardware refresh and aim to keep them up to date with software deployment systems, the reality is that once users start using a machine, they almost instantly make the machine unique to them. Either through changes to configuration or the introduction of applications and plug-ins, the machine strays away from the gold image. This limits IT’s ability to manage the machine because when a user reports a problem, it is difficult to know whether the cause is a real problem, a failure in application deployment, or a user-introduced fault. Consequently, much time is wasted trying to understand where the fault lies and, frequently, if a cause cannot be quickly located, the machine will be re-imaged. This wastes IT time and disrupts users from working. Additionally, most users’ machines are defective in some way but users wait for major problems before calling the help desk, hence the low quality of service of typical machines.

The root cause is that we have long had an implicit understanding that a PC consists of an operating system, applications and the user environment but that the tools we had to manage PCs could not manage each component as effectively as they should because each component interacted with the others. For instance, installing an application frequently changes dlls in the operating system and can change the way other applications work. Of course, user changes can throw the system another curve ball. The use of virtualization technologies such as operating system virtualization, application virtualization and user virtualization allows us to keep each of the components separate, allowing us to manage each effectively.

This is the problem that vendors such as Citrix, VMware, Microsoft and AppSense have worked on. The solution ushers in a new dawn for desktop virtualization. Essentially, each of the components of the PCs software image can be delivered on demand in such a way that the software is always at the most recent standard. This makes the platform very predictable and economic to manage. Similarly, the user environment is delivered on demand but represents all the user-specific aspects of the machine. In combination, OS virtualization, application virtualization and AppSense’s user virtualization creates a platform that is far easier to manage, yet gives users a familiar and productive PC experience. It is this combination of characteristics that has led to so many organizations trialing desktop virtualization. What is more important is that the success of those tests is now leading to broader deployments during 2010. One of the side effects of those broader deployments will be to validate the economics of desktop virtualization at scale.

Unlike the flying car, desktop virtualization has become economic to deploy and acceptable to users. During 2010, we will see larger deployments being successfully rolled out.