Microsoft Downplays Desktop Virtualization

Tech Analysis: In a touring workshop to promote Windows 7, Microsoft is calling VDI, or virtual desktop infrastructure, a costly venture inappropriate for larger enterprises. That's not surprising, given Microsoft's vested interest in having fat clients continue to run on PCs, and IT administrators should not let that deter them from investigating VDI. As eWEEK Labs analyst Cameron Sturdevant points out, there are positives and negatives to VDI environments, and virtualization of desktops is very different from server virtualization.
##CONTINUE##
Microsoft has been taking a "Windows and the Enterprise" workshop on tour to promote the future Windows 7 operating system. One of the key messages from the workshop: Virtual desktop infrastructure is an expensive proposition unsuited to large implementations.

Microsoft has a vested interest in the continued use of fat clients running on PCs, and thus is an unreliable source when it comes to this topic. During a recent presentation in San Francisco, company officials said Microsoft VDI might be suitable for implementations of up to 100 seats, but no more. Microsoft does partner with Citrix Systems for large-scale implementations that also include VDI.

Now, there are some valid reasons to say no to VDI today. As I pointed out in January, the computing back end needed to support VDI is considerable. Local resources, including storage and the CPU, must now be provided in the data center. Reliable network connectivity and capacity take on a whole new level of importance when dishing up entire desktops along with the accompanying applications.

Click here to read more about Microsoft's virtualization efforts with Citrix.

There are important characteristics of VDI that make it substantially different from server virtualization—the current golden child of IT cost savings. Desktop workloads involve far more applications than are typically found on virtualized servers. These applications are also combined in much greater variety on desktops than on servers. Finally, humans interact with desktop workloads in a way that is unheard of with data center servers. Users elicit all sorts of erroneous behavior from desktops, often in the name of, "Just trying to get my work done."

The variety of workloads and this large amount of human interaction combine to pose a huge challenge that will not easily be overcome. However, there are technological choices and advances being made today that make the challenge significantly smaller. I am suspicious of one of them—the thin client—because its arrival on the threshold of large-scale adoption has been heralded more than a few times. The other is of a more practical nature: changes to the RDP (Remote Desktop Protocol) that make it more suited to handling bidirectional audio and video, which add telephone services to thin clients.

I've seen several demonstrations—including at the Windows and the Enterprise workshop—of audio and video capabilities that take a big step in the direction of enhancing virtual desktop technology. Some of the advances made in bidirectional audio and video are likely the precursors for supporting CAD/CAM applications in a virtual desktop environment, something VDI is entirely unsuited for today.

There are some valid points in favor of saying yes to VDI, or—more likely at this point—contemplating a virtual desktop offering. Centralized control over the desktop is among the top features that draw me to this technology. Putting control of desktop features and applications in the hands of business managers—please note that I did not say "end users" or "IT staff"—is made possible in a cost-effective way with VDI. Business managers with the able assistance of IT staff can decide what tools to make available to end users with which to accomplish their work.

These tools—let's call them applications and services—can then be deployed, all properly configured and in good working order at the start of work. VDI provides a strategic platform for massive deduplication of data. It also provides a much more secure work environment by preventing users from making desktop changes that allow malware to strike, such as installing rogue software.

The thing that seems to impede Microsoft's ability to see VDI as a wide-scale solution is that it questions the need for Windows on the end-user system. This shouldn't stand in the way of IT managers and business leaders interested in VDI. The possible benefits in terms of better regulatory compliance through strong enforcement of desktop and application configuration, tighter security through frequent desktop refreshes, and tighter lockdown are compelling reason to question whether the days of the fat client are numbered.

-----------------------------
BY Cameron Sturdevant
Source：eWEEK.COM

Copyright ©1996-2009 Ziff Davis Enterprise Holdings Inc. All Rights Reserved. eWEEK and Spencer F. Katt are trademarks of Ziff Davis Enterprise Holdings, Inc.