Recognize the Real Promise of Hosted Desktop Virtualization
One recently introduced technology whose adoption trajectory seems to be following that of a typical hype cycle is hosted desktop virtualization. Hosted desktop virtualization (sometimes referred to as virtual desktop infrastructure, or VDI) represents a completely new approach to the deployment of enterprise desktop computing. This article will outline the genesis of hosted desktop environments, and explain why today's technologies are able to fully deliver the benefits that have been promised in the past. In order to understand what makes it so different, let's first take a look at desktop computing as it typically exists today.
The traditional model for deploying desktop computing within an enterprise involves the use of personal computers (PCs). Each of those PCs has a fully-featured operating system installed on it, which provides all of the basic computing operations and functions. Applications such as word processors, spreadsheet programs, and web browsers are then installed on top of that operating system in order to provide users with the tools that they need in order to perform their jobs. All of the data associated with the operating system and the applications is stored on the internal hard drive of the PC itself. To provide some real-world context, I've written this article using a traditional desktop approach: I used a word processing application, OpenOffice Writer, which was running on top of the Fedora 11 Linux operating system, which was in turn running on my desktop computer. All of the processing and data required for OpenOffice and Fedora was resident on my computer, as was the file representing the article itself.
In hosted desktop virtualization, the operating system and applications that make up a user's desktop environment actually reside on servers located in a centralized data center. Users connect to these desktop environments, known as virtual desktops, using a "thin client" that has a minimal amount of software installed on it and a network connection to the data center. Unlike with traditional desktop deployments, no data is stored on these thin clients. So, getting back to our real world example...if I were to write this article using hosted desktop virtualization, my application (OpenOffice), along with my operating system (Fedora 11) and the file representing the article itself, would all reside on a server in one of my organization's data centers. There it could easily be secured, backed-up, and restored if needed.
Sounds pretty good, doesn't it? Now, we need to acknowledge the fact that the idea of thin client computing is nothing new. Like so many other things in IT, what's old is new again. Thin clients represent a return to many of the same principles that were in place during the days of the so-called dumb terminal, as well as the thinking that was behind X terminals and later the Network Computer promoted unsuccessfully by Oracle Corporation. But in reality these earlier attempts at thin client computing could not meet the needs of enterprise users, and as a result the PC remained the dominant choice for enterprise desktop deployments. Along the way, operating systems developed larger footprints and got more complicated to manage, applications became more complex, and data began leaving the office as workforces became increasingly mobile.
