Microsoft's Virtualization Strategy Coming Into Focus

Will Microsoft's server virtualization push change the rules of the game?

Given the dominance of VMWare Inc. in the x86 virtualization space, Microsoft Corp.'s own efforts -- which until recently have seemed either fitful or faltering -- have fallen curiously flat.

That's set to change with the forthcoming release of Windows Server 2008 and Microsoft's long-promised Hyper-V virtualization technology -- although just how much it will change is anyone's guess.

Things became more clearly focused in December, when Microsoft unveiled its first Hyper-V public beta. On top of this, Redmond announced that it's already developing the next version of its System Center Virtual Machine Manager (SCVMM), which allows customers to quickly provision and configure new virtual machines and centrally manage virtual infrastructures running on Hyper-V, Microsoft Virtual Server 2005 R2, VMware ESX Server and Virtual Infrastructure 3 (V13). (The first version of SCVMM shipped in October 2007.)

Microsoft confirmed that the final version of Hyper-V will ship within 180 days of Windows Server 2008's RTM date.

How is Microsoft's virtualization story shaping up? Nicely, industry watchers say -- although Redmond can't realistically hope to challenge, much less topple, the VMWare juggernaut. In fact, VMWare -- which has a near-unassailable position in the market -- might currently be the least of Microsoft's concerns.

For one thing, says industry veteran Charles King, a principal with consultancy Pund-IT, the x86 virtualization segment is, well, virtually teeming with competitors. "It is hard to think of a commercial technology that evolved more substantially during 2007 than x86 virtualization. The year included a stellar initial public offering (IPO) from market-leader VMware, new and/or enhanced offerings from other specialty players, like Virtual Iron and SW Soft [now Parallels], the high-profile acquisition of XenSource by Citrix, and new competitive offerings from major vendors, including Dell and Oracle," he points out.

That being said, King allows, Microsoft is, after all, Microsoft. "[M]any believe … Microsoft's Hyper-V [formerly code-named Viridian] product the most likely to shake up and substantially change the virtualization market," he says. "By its sheer weight, Microsoft can potentially alter most any competitive landscape it chooses. Microsoft has also proven it knows how to play the long game when it deems the eventual goal worthy of substantial patience and investment."

With Hyper-V, then, Microsoft appears to be digging in for the long term. There's a lot to like in that product, King concedes, but -- as with many of Redmond's more ambitious efforts -- Microsoft has had to significantly retrench with respect to what it actually plans to deliver with Hyper-V.

"The company's beta release of Hyper-V allows Microsoft to end the year on something of a high note, though the product is not quite what the company originally said it would deliver," he says. "Most glaring is a notable reduction in the number of processor cores supported [down to just 16 from Microsoft's previous promise of 64] and the elimination of Live Migration, a feature aimed to compete directly with VMware's VMotion."

Disappointing, yes; but not crushing, according to King. "These changes are troubling and will likely disappoint some larger customers, but Microsoft scores points for its larger strategic aim," he indicates. "The company is essentially betting that virtualization works better and provides more value as an essential, affordable part of the server operating system than as a standalone application."

The question, King concedes, is whether the market is ready for Microsoft's approach -- especially given the non-operating-system-centric approaches of VMWare and other competitors. Microsoft has hedged its bets in this regard -- it now positions Hyper-V as a standalone hypervisor, which is a testament to the strength of VMWare and other competitors in the virtualization segment -- but it still wants to shop its operating system-centric take on virtualization (see

"Few argue the effectiveness of virtualization as a tool for server and application consolidations and for enhancing server availability and business continuity, but some do complain about the price of standalone commercial solutions," King points out. "Offering Hyper-V as an inexpensive option for Windows Server 2008 should aid Microsoft in driving a larger message about virtualization's essential strategic role across numerous IT and business processes."

In spite of its many positive attributes, King doesn't believe Hyper-V, in its first iteration, anyway, will be a slam-dunk success -- although it should prove attractive in Windows shops.

"Does that mean that Hyper-V will be an instant success? Maybe, but maybe not. For companies planning to deploy Windows Server 2008, Hyper-V qualifies as a miniscule additional investment that can be easily uninstalled if it fails to work out," he concludes.

"At the same time, despite its virtualization ambitions, Microsoft is entering a youthful but rapidly maturing market that sports far better-known and more sophisticated competing solutions such as VMware's ESX Server. In addition, while the low cost of Hyper-V will make it attractive for many businesses to test, how many of those evaluations will result in substantial deployments is difficult to say, particularly among larger organizations currently using VMware's VMotion solutions."