2012 promises to be a year of monumental upheaval in the virtualization technology world, so hang on to your hats: The ride is going to get very bumpy! The reason? 2012 will likely mark the end of VMware server virtualization hegemony.
The company has been the undisputed leader in server virtualization for as long as most people care to remember because its products have been far more advanced than anything any other vendor could offer and because of the comfort factor. VMware's products may be wildly overpriced, but no one ever got fired for buying VMware.
But plain vanilla server virtualization is now a commodity, and both Citrix and Microsoft -- and even Red Hat -- will offer very powerful alternatives in 2012 at a fraction of the price. That means hypervisor heterogeneity is going to become the order of the day, with an increasing number of companies looking to move at least some of their virtualized workloads from their pricey VMware infrastructure to an alternative hypervisor that can do the same job for less money. And who can blame them?
Of course, once heterogeneous virtualization environments become established, it will be much harder for VMware to make headway. That's because each time a company considers virtualizing new or existing workloads there will be a kind of virtualization triage: Does this workload have requirements that merit putting it onto the more costly VMware infrastructure, or will an alternative hypervisor be up to the job? Increasingly, companies will realize VMware is not required.
Which virtualization system will emerge as the biggest competitor to VMware in 2012? It's a tricky question to answer, but my bet is on Hyper-V. That's not to say Citrix's XenServer 6 is not a creditable challenger, but Hyper-V 3, due out some time this year, looks very full-featured and, well, it's from Microsoft. That means it's going to get an almighty push, and many Microsoft shops have the skills and infrastructure in place to take advantage of it with a minimum of disruption -- particularly in the SMB space, which will be responsible for a large part of the growth in server virtualization.
With heterogeneity comes the problem of management: No one really wants to manage their two (or more) virtualized server estates separately, but the management platforms offered by hypervisor vendors are rather limited in their scope for looking after other vendors' hypervisors. That leaves a huge opportunity for third-party management software makers to offer solutions that either manage multiple hypervisors, or enable existing management platforms to manage additional hypervisors more effectively. The only question is how long these companies are able to remain independent before being snapped up by the hypervisor vendors themselves. It's a good bet they will survive 2012, but further out it's hard to see the most successful ones staying independent for long.
So is it all gloom and doom for VMware? The company may be about to lose a hefty slice of the enterprise virtualization market to Microsoft, but it looks like it will be pushing hard in the cloud service provider space -- a market likely to grow rapidly. 2012 may well be the year that VMware really establishes itself in the public cloud. That's not to say Citrix and Microsoft will not be pushing hard in this area too, but VMware is probably two to three years ahead: Don't expect Microsoft and Citrix to catch up until 2014 at the earliest.
What about a long-shot for 2012? No one seriously believes this is the year desktop virtualization finally makes the big long-predicted breakthrough that never quite seems to happen, but Citrix's SMB-oriented VDI-in-a-Box (acquired with Kaviza) might be the surprise success for 2012. It's easy and inexpensive, and it might be just what a lot of companies are looking for. Watch that space!
Paul Rubens is a journalist based in Marlow on Thames, England. He has been programming, tinkering and generally sitting in front of computer screens since his first encounter with a DEC PDP-11 in 1979.