With a little more than a week to go in 2008, I found my mind drifting to the events of this year -- at least those relevant to IT. It may seem obvious, but I think that we'll look at 2008 as the year that virtualization finally landed. After more than a decade of x86 virtualization, it's to the point now where it's not only a viable technology, but it's being adopted in every IT sector, from production to development, from small to large business.
It's no secret that virtualization will provide the framework for the next generation of IT infrastructures, but it may be that this was the year that that became a firm conclusion, not a potential.
VMware still leads the pack, and VI3 is still the flagship product for enterprise virtualization, but VDI made big news this year too, in the form of Citrix taking the Xen plunge and planning for the next decade -- a move that should turn out very well for the company, assuming they play their cards right. VMware's Thinstall acquisition is bearing fruit, and the panoply of other players in the VDI/broker space, like Pano Logic, are pushing all the right buttons.
I'm hearing more and more interest in virtualization in general and VDI in particular. The onerous requirements of Microsoft Vista and the aging XP deployments in most organizations, coupled with the economic downturn seem to be pushing this fairly hard. The three-year desktop cycle is being lengthened due to budgetary stress, and VDI would seem to fit nicely into that space.
This is both good and bad for hardware vendors. With the profit margins (and some might say the quality) of corporate desktops falling through the floor, replacing those units with a thin client connecting to a VDI implementation looks more and more reasonable. This obviously means that more organizations will be stretching their existing desktops, and even turning them into thin clients with any number of tools. The good news for hardware vendors is that the higher-margin server space will be opening up a bit. Yes, virtualization can mean that fewer physical servers are necessary in many infrastructures, but VDI will necessarily increase the number of virtualization hosts, which may or may not bridge that gap.
In other news, blades may have finally found their place in the world. Blades have been around seemingly forever, but outside of some specialty shops and HPC work, they never really made sense for many medium to large enterprises. With virtualization, they make much more sense, especially with the all but required homogeneity of virtualization farm servers.
Microsoft's getting serious about virtualization, even if its current platform isn't enterprise-grade. If history is any guide, the company will have something in the near future that can compete with VMware, a scenario that will hopefully be good for IT as a whole. As it stand now, however, Microsoft selling Hyper-V as an appropriate solution for mid- to large-infrastructure virtualization is simply a non-starter.
So what might 2009 bring for virtualization? Hopefully a solidification of virtualization standards, such that VMs built for any hypervisor can boot from any other. That's a long shot, but it would be nice. Advances in memory allocation and sharing, specifically for VDI workloads, are already in the works, and perhaps 2009 will be the year that VDI hits all sectors of the market.
As with everything in IT, 2008 was a hectic year, and 2009 promises more of the same. After all, it is IT.