Microsoft, the big absentee of virtualization

Where is Microsoft while VMware is delivering virtualization innovations one after another, driving the whole IT market quite all alone?

Let’s consider the Virtual PC / Virtual Server development cycle from 2002, when the Redmond giant acquired the VMware competitor called Connectix, to today:

  • 2003 – Virtual PC 2004
  • 2004 – Virtual Server 2005
  • 2004 – Virtual PC 2004 Service Pack 1
  • 2004 – Virtual Server Migration Toolkit
  • 2005 – Virtual Server 2005 R2 (originally planned as Service Pack 1)

In 4 years customers saw few innovations (some readers will remember that Connectix was already working on a server product when Microsoft acquired them), no enterprise management tools, no serious physical to virtual (P2V) migration tools.

Now, avoding any consideration on the early strategy, we have to ask why Microsoft is approaching virtualization in this way since VMware started becoming a serious competitor (probably after being acquired by EMC) and monopolizing the market.

3 possible reasons:

  • Microsoft doesn’t really see VMware as a serious competitor and waited VMware success to see how the market would answer to pervasive virtualization
  • Since Connectix acquisition Microsoft changed its mind and is still doubtful on what to do
  • Microsoft is concentrating all efforts on the announced Windows Hypervisor (codename Viridian)

Considering (and hoping for) the last hypothesis we have to face a new issue: Microsoft Windows Vista (the client operating system to succeed XP) has just been delayed to end 2006 (and worldwide distribution on January 2007).
And this is the biggest issue considering Microsoft has announced Windows codename Longhorn Server (the server operationg system to succeed 2003) to be released 6-8 months after Vista, and Windows Hypervisor to be released 1 year or more after Longhorn.

So whay could happen to virtualization strategy now that Vista scrambled all marketing plans?

  • Microsoft will gain unexpected time by Vista delay and will deliver Windows Hypervisor immediately, within Longhorn
    (this could be highly unlikely since Microsoft could need as much developers as possible on Vista to respect the new deadline)
  • Microsoft will delay Longhorn and Windows Hypervisor accordingly to Vista release, unchanging distances between products releases

Let’s consider once again the last hypothesis: virtualization customers, now waiting for Virtual Server 2005 R2 Service Pack 1, would stay without technology enhancements for another 2-3 years.
But this would be completely disruptive for the Microsoft business model. So I would expect a fast plan rearrangement with another, unexpected, minor release of Virtual Server, before Windows Hypervisor will be out.

The real problem is: what Microsoft will be able to offer in this scenario against VMware and Xen (which is now highly supported by Red Hat and Novell efforts)?

Note: although I’m a Microsoft Most Valuable Professional (MVP), no reserved informations have been disclosed in this article.