Written By: Bernd Harzog

In Star Trek TNG (The Next Generation), fighting the Borg was futile because you always lost. Is Hyper-V the “Borg” of virtualization technologies, destined to “assimilate” all that stands in its way? Let’s take a look at Microsoft’s playbooks from years (and wars) past to glean some insights as to how this war (and it is now a war) is likely to turn out.

Choosing the Correct Precedent
It is easy to say that Microsoft vanquished Netscape by giving away what Netscape was trying to sell, and therefore won. But, the browser is logically an extension of the operating system (what OS on what platform does not include a browser these days?), so using this as a precedent is like saying that just because spell checkers became part of word processors, means that Microsoft is destined to obliterate VMware.
 
I think that SQL Server is a much more illustrative precedent than is the browser, the spell checker, the spreadsheet, personal productivity suites (Office, SmartSuite, etc.), and email (Outlook, Notes). The reason is that there are a wide disparity of requirements between very high end business critical applications (of which there are relatively few per company), and the far more numerous important applications that are not “critical”. Critical in this context means that it must be up and running virtually 100% of the time, with good performance to all users. An electronic trading application at a financial institution is business critical – an outage of a second during the trading day can cost millions of dollars. Email is extremely important, but if it stops working for a couple of minutes every now and then this is for the most part tolerable. If you recall, when Microsoft entered the relational database market with SQL Server, the dominant products were Oracle, FoxPro, and dBase. Microsoft licensed the RDBMS technology from Sybase, bought Foxpro, and proceeded to make SQL Server into a database that was good enough and easy enough to use for all of the companies that were stepping up from FoxPro/dBase to something that could really handle multi-user database applications. Soon SQL Server became the volume leader in RDBMS shipments. Soon more DBA’s knew SQL Server than Oracle. Soon, it was much easier to build a SQL Server application with Microsoft development tools than it was to build one to Oracle. Soon, there were many more solution providers who understood SQL Server than Oracle. Finally, it was just so much easier for a company that did not need the upper 5% of scalability, availability and management to use SQL Server than it was to use Oracle. At this point does Microsoft care that the 5% of the most demanding database applications in world use Oracle and that everything else could use SQL Server (and use it for a lower cost of ownership than Oracle). I do not think so. So, Microsoft won.
 
Here Is What to Expect (the most likely case)
If you think that the SQL Server example above is illustrative as to what to expect, then here is what I think you should expect in the virtualization market:

1. Microsoft will capture the high volume low price segment of the market by commoditizing it. Commoditizing a market means that you reduce perceived differentiation to zero, so that customers buy whatever is easiest and cheapest to buy and implement. Microsoft will make Hyper-V good enough, easy enough and cheap enough for the 95% of the world that has not done anything with virtualization to date.
2. Hyper-V will make virtualization accessible to a set of customers that have not even considered VMware to date. This will principally be driven by the fact that while many high end enterprises and VARs have people who understand a “Unix” oriented hypervisor underneath their Windows Servers, there are a hundreds of thousands of end user companies and VARS for that do not have that skill set but  that do understand “Microsoft”.
3. The shipments and installations of Hyper-V, will quickly outstrip anything and everything that VMware has done to date and will do. Expect to read reports in the press in the next 18 months that Microsoft has become the “market share leader” in virtualization platforms. VMware aficionados will scream that “Microsoft is not enterprise ready”, or is it “not manageable” and they will be right. But it will be good enough out of the box for 80% of the new virtualization projects, and will grow over time to become good enough for 90% to 95% of those projects.
4. Large enterprises that already have adopted VMware will be driven by cost considerations to segment their virtualization projects into ones that “need” HA, DRS, VMotion, Disaster Recovery, Applications Response Time Monitoring (B-hive), and those that do not. They will find that 20% of their high end business critical applications require these high end capabilities and that the other 80% do not.
5. In a bizarre twist, many enterprises will find that the “low hanging fruit” that they initially virtualized does not need the high end features of VMware, and will take what they already have and move it to Hyper-V just because it will be less expensive. However, those same enterprises will find that they need VMware for the next set of more business critical applications that they have not virtualized yet.
6. Microsoft and Citrix will win both the early market and the long term market for virtualized desktops. This market plays to the strengths of the Microsoft/Citrix partnership, and the strength of these two companies together will overwhelm anything that VMware can come up with in VDI.
7. #5 and #6 above will create a “Perfect Storm” of projects for VAR’s and solutions providers. 80% of what is on VMware today will be a candidate for migration to Hyper-V, while the business critical applications that have not been virtualized yet in large enterprises, will be perfect candidates for the next VMware projects. VDI projects will be much more complex (because they involve users) and therefore profitable for the VARS, and VARS will leverage their existing relationships with Citrix on this front.
8. Microsoft will fight a multi-front war against VMware. One front will be management where the strategy will be “embrace, extend, and extinguish”. This is already visible in Systems Center Operations Manager Virtual Machine Manager (SCOMVVM) where Microsoft has built in the ability to manage both Hyper-V and VMware ESX environments into their product. Microsoft will also:

•  Leverage its ownership of Windows to make sure that Hyper-V is the best hypervisor for Windows Guests (again – where is the volume market?)
•  Leverage its development tools to help developers (one of Microsoft’s core constituencies) build virtualization aware applications
•  Leverage its training network to ensure that MSCE’s spin up on Hyper-V
•  Leverage its relationship with thousands of software vendors to certify their applications on Hyper-V,
•  Leverage its relationships with ten’s of thousands of channel partners and VARs to make sure that they all know how to sell and install Hyper-V
•  Leverage its world class ability to partner with companies that fill in the holes in Hyper-V with things that are needed by different sets of customers.

Using SQL Server and Oracle as the example, if VMware ends up being the Oracle RDBMS of virtualization, is that a bad end? Oracle was able to maintain premium pricing for its RDBMS product (and still does so today) despite the onslaught from Microsoft SQL Server. If VMware plays its cards correctly, and continues to add the right pieces of differentiating value to its high end customers and VARS, then VMware should be able to continue to be the high end virtualization platform of choice. So, VMware will continue to prosper and grow because most really business critical applications have not been virtualized yet. In other words, this is not by any stretch of the imagination of “zero sum game” where just because Microsoft wins, VMware has to lose. Under this scenario, in fact, both companies stand to win and win big over the next 5 to 10 years.
 
Could VMware Fail Miserably?
Yes, any company is capable of failing miserably. Contrary to popular belief this most often does not occur because they were done in by a competitor (in this case Microsoft), but because of gross stupidity and lack of foresight on the part of the management team of the company in question. So what would be some grossly stupid things that VMware could do (or not do) that would be tantamount to VMware digging its own grave? Here is my list:

1. Arrogance. “We have 80% of this market”, “This market is ours”, “They (Microsoft) do not understand this market”, “We have our customers locked in to our technology – it would be too expensive for them to switch”, “We have the channel locked up”, “We have a three year lead in product functionality”, etc. Market leaders who dominate early markets (as VMware has) tend to both be arrogant, and not realize they are being so. This is a great way to fail.
2. Alienating valuable friends in the channel. I have spoken to many VARs and Service Providers who grit their teeth all of the way to the Bank in their relationship with VMware. As long as they have no choice they will continue to endure pain to capture the VMware project revenue flow. But now those same partners can choose to do Hyper-V projects or Citrix Xen projects and not deal with some of the painful aspects of doing business with VMware. VMware has become known for being difficult for channel partners to deal with, and for competing with their channel on the fronts of both license revenue and large services projects. Both Citrix (who has a large and loyal VAR community that largely overlaps with the VMware VARs) and Microsoft will take great pains to be more channel friendly than VMware. VMware is going to have to grow up and realize that it needs the channel more than the channel needs it. If VMware does not grow in this regard, it will ultimately die as a result.
3. Alienating valuable friends in the partner software community. One of Microsoft’s fundamental assets is it relationships with developers in companies that build add-on software products to its operating systems. Microsoft is extremely adept at being very accommodating to vendors that add value, even if they compete with some aspect of the Microsoft platform. This broad view of who is a friend (anyone who writes to the Windows API’s) is a huge part of the Microsoft culture and ecosystem. By way of example, Microsoft can count Citrix as a partner on the virtualization front, as both companies are working together to make sure that Citrix’s Virtual Desktop offerings are a clean value add to the Microsoft Virtual Server offerings. If VMware starts to shut down sales help to vendors in the management space that “support” their platform, but that “compete” with some feature of the broader VMware product set then we will know that they have gotten it wrong.
4. Inflexible and unresponsive pricing. VMware needs to realize that it is not immune to the fundamental economics of IT Operations. Those fundamentals are based upon the imperatives of doing more with less budget every year than the year before. VMware has been largely isolated from the brutal economic pressures to date because it has been the only game in town, and because it has offered a compelling hard dollar ROI. But now that the same “return” will be available with less “investment” from Microsoft and Citrix, VMware will have to adjust. This does not mean that VMware will have to meet Microsoft and Citrix in terms of pricing, but it does mean that VMware will have to come close to competing on a head-to-head basis for the commoditized parts of its product line (the hypervisor), and focus upon charging premium prices where it has true product differentiation (the management pieces). In other words, if VMware does not restructure its pricing in the second half of this year, it will be sowing the seeds of ultimate failure due to pricing arrogance.

Note. I do not think that VMware is stupid enough to make any or all of the mistakes listed above. But if they do – look out.

Could VMware Stem the Tide?
Is there way for VMware to continue to own the high end market (which is critical to its survival) and also become the market share leader in the volume markets being targeted by Microsoft and Citrix? Yes, there is a way. It would require that VMware focus upon market penetration at the expense of revenue and profitability. In other words it would require that EMC (VMware’s parent) allow VMware to run at break even levels or even a loss for several years. It would require that EMC use the profits from its other businesses to cross-subsidize market share expansion on the part of VMware. This is exactly what Microsoft and Citrix are doing in this market. Microsoft is using profits from Windows and Office to fund its market share expansion in virtualization. Citrix is using its profits from Presentation Server (now called XenApp) do fund its growth in desktop virtualization. This is a business decision that EMC and VMware need to make at the Board of Directors and Management team level. My guess is that given the choice between growing profits with a high end strategy for the next 10 years, and no profits for the next 5 years with a market penetration strategy, executive management at EMC and VMware will choose to deliver profits instead of promises to their shareholders. I cannot think of any examples of executives of publically held corporations in American making this kind of a decision in the last couple of years, so unless a hedge funds buys EMC or VMware, the numbers basically say that this is not going to happen.
 
What Does This Mean For Management Vendors?
Management vendors of all stripes face enormous challenges and opportunities. I think a year from now almost all VMware customers will also have Hyper-V. Microsoft is betting on this by having a cross-platform offering with SCOMVMM. Any independent management software vendor that wants to prosper in this situation, needs to also be cross-platform (support VMware, Microsoft and Citrix), and most importantly clearly add value to the disparate management tools of the platform vendors. Management vendors will need to both be consistent across platforms (give customers one way to do things), while also taking advantage of unique features in each platform.  Customers and VARs should actively look at third party management solutions across the entire range of management needs since it is unlikely that the platform vendors are going to address most of the needs.

  

• DABCC User:
• Guest
• Subject:
• Message:
• Image Code:
• 

Latest Articles