My comrade-in-Labs Cameron Sturdevant is writing about the lukewarm attitude that Microsoft appears to have toward virtual desktop infrastructure. To be sure, we at eWEEK Labs have our reservations about VDI, or virtual desktop infrastructure.
For one thing, VDI, as it's now implemented, isn't positioned to deliver the same dramatic cost savings as can its more handsome sibling, server virtualization. For another, I don't like the idea of leaving my desktop environment to the mercy of an Internet link.
Microsoft, however, appears to be giving the concept somewhat shorter shrift than it is due, perhaps because VDI clashes somewhat with the company's fat (or rich, if you prefer) client business model. If desktops begin to migrate to the server room, they'll be migrating to a place where Microsoft wields significantly less control than they do in the desktop and notebook realm.
This clash of desktop business and deployment models calls to mind Red Hat's stance toward the desktop as a commercial offering. As the company's CEO Jim Whitehurst explained to me during our interview last September, Red Hat officials have demurred at pushing the desktop as a product because they haven't hit upon a model for monetizing it. Instead, Red Hat has situated its continuing desktop efforts in the incubator of the freely available, community-supported Fedora distribution.
However, now that it has begun digesting the KVM virtualization and SPICE desktop presentation technologies it acquired through the purchase of VDI provider Qumranet, Red Hat seems to be zeroing in on a way both to monetize the desktop and continue to give it away. Far from threatening current income streams, Red Hat can mine a whole new revenue vein by selling and supporting the tools required to deploy, host and manage VDI.
As a result of Red Hat's recent agreement with Microsoft to support each others' operating systems on the companies' respective virtualization platforms, Red Hat will be able to serve up Windows desktop sessions from its own Linux-based platform.
Considering that the VDI market is already populated by well-established vendors, such as VMware and Citrix, it's true that Red Hat will have some catching up to do. What may end up propelling Red Hat beyond those more deeply rooted competitors is that, unlike Citrix and VMware, it has a high quality desktop offering of its own waiting in the wings.
With the addition of VDI hosting and management revenues, Red Hat should be able to offer up a more stable free desktop than they've previously made available, something along the lines of a Fedora Long Term Support version (to take a page out of Ubuntu's playbook), and offer customers the option of shifting some of their desktop sessions from Windows to Linux.
By pairing VDI with a free or low-cost client OS, Red Hat could boost the cost savings that companies can expect from VDI, thereby addressing the first, cost-based set of VDI reservations that I expressed above.
More importantly, between the work that Red Hat has done around stateless Linux, Fedora on a (USB) stick and client-side hypervisor implementations, it seems to me that the company has built a compelling foundation for addressing the second set of VDI concerns I mentioned-those related to online/offline desktop accessibility.
Canonical and Novell have made some real strides toward taking the Linux desktop mainstream, but I've long believed that until the Linux world's biggest player resolves to tackle the desktop, the dawn of desktop Linux will remain deferred. As this new VDI chapter unfolds, I'll be watching closely to see if Red Hat has indeed figured out a way to make the desktop pay. ??
Executive Editor Jason Brooks can be reached at email@example.com.