Home > Uncategorized > A Hidden Gem VMware Should Invest More For Its Long Term

A Hidden Gem VMware Should Invest More For Its Long Term

For a long time, VMware wanted to grow upward the stack but did not work out well. Instead, it took the horizontal strategy as I discussed in my blog article two years ago. So far, the new strategy has worked well.

Although VMware has its strong hold in enterprise because it is the company that created the virtualization market by itself, the risk of being challenged by other vendors is far from over. Technically speaking, the touch point for the VMware products is so small that customers can move away from VMware without too much effort. That is what has been happening in SMB market with Microsoft Hyper-V, and in enterprise market with open source KVM.

Time to learn how to "Google" and manage your VMware and clouds in a fast and secure


Because the overall market is still growing, all the vendors still see increases in revenues. At a future point of time, the market will mature and saturate. By then, any loss in market shares would mean revenue decrease. Every vendor has to prepare for that point and come up with a strategy for preventing loss of market share. That is especially important for VMware because it has been giving market shares to the competitors like Microsoft and KVM over the last few years even though it’s still #1 today.

There are many different approaches to prevent customers from leaving for other products.

On business side, a company can lower its price to make the product more competitive compared with other vendors. But with Microsoft Hyper-V, which comes free (not really free, but is free if you have Windows license already) with Windows server license purchase, and open source KVM, it’s not really a feasible approach unless VMware turns itself into an open source company as well. Another approach on business side is to have a strong ecosystem around core products. VMware has been doing pretty well on this side. On track record, Microsoft is a real master on building partner ecosystems – just think how it did for its Windows platform. Given VMware’s lead in time to market, its vSphere will continues to be de facto virtualization platform over the years, but more and more third parties have started or will start to support Hyper-V.

On the technical side, VMware had wasted a few years in attempts to grow vertically upward the stack. Its technical advantages have been narrowed down quite a bit to a point where customers don’t care about the differences either they don’t need some new features, or the benefits from the features are really marginal. VMware has to innovate more on horizontal strategies. Among many different possibilities there, there is a low hanging food that would generate biggest benefit with lowest investment.

This is an existing product – VMware Tools.

Why it’s important? Besides its existing features, there are quite a lot of things VMware can offer with this tool. The goal is to make the workload sticky even though virtual machine is not sticky. At one time, people thought VMware would buy Suse Linux, but it did not. I think it’s a right decision. But it does not mean VMware should not do more in OS to make its virtual machine stickier. I think the answer is the VMware Tools.

As we know, the VMware Tools is installed in OSes that vSphere or Workstation/Player supports. It helps provide better network drivers, mouse cursor experience. In short, VMware Tools makes a virtual machine running faster/better on vSphere.

While there are already many features in VMware Tools already, VMware should innovate more on it. For example, it can provide feature to dynamically construct virtual machine from base virtual machine template and provisioning scripts like Chef, guest OS level automation, etc. The possibility there is endless.

As always, there is no way for VMware to make money out of VMware Tools. Not before, neither in the future. But this is strategically important for the company to keep its platform stickier.

Categories: Uncategorized Tags:
  1. May 21st, 2014 at 01:14 | #1

    Steve, I see where you’re going with this but I wonder if that opportunity is about to pass them by all together. Puppet and Chef are starting to reach enough maturity that implementation can be done effectively without the influence or interaction of a VMware administrator. Additionally, Microsoft now has Desired State Configuration for Windows systems and as announced recently can do desired state on Linux systems (in tech preview). To my chagrin, and no surprise, this functionality isn’t fully baked. In the end though it seems that some simple setup in the VMware template can allow VMware administrators to be hands off all together except to fire up the template occasionally, let it get its updated configuration from centralized management, and convert it back to template. This capability is hypervisor/platform agnostic.

    I think the bigger value add using VMware Tools would be in increasing capabilities to manage the virtual machines after deployment and on tasks that may be outside of Puppet/Chef/PowerShellDSC native capabilities. VMware tools allows administrators greater abilities to do things to the VMs that are not network access dependent. PowerCLI already has ability to inject scripts and perform actions within the guest OS via the hypervisor and VMware tools. Perhaps expanding these capabilities in PowerCLI or vCO to provide more out of the box functionality to vCAC would be more the way to go. Additionally providing integration with Puppet/Chef that can communicate via VMware Tools instead of the network may prove to be more efficient at scale. You show an organization that they can cut their desired state implementation in half via VMware tools communication in lieu of network communication then they may be on to something unique.

    I like the thought. Very interested in more of your thoughts on how VMware tools could be more leveraged. They already enable quite a bit of functionality as listed at http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=340 I do wonder how much of that is wholly unique at this point, much like the hypervisor and management features.

  2. May 21st, 2014 at 10:19 | #2

    Hi Josh,

    Thanks a lot for sharing your thoughts. I like your idea of integration of Puppet/Chef. Because VMware Tools is an agent itself, and administrators can save time by avoiding installing and managing another agent of Puppet/Chef. Having one less thing to install and manage is a plus. As you pointed out, the Puppet/Chef is network dependent. In other words, they will NOT work if there is no network connection configured in the virtual machine.

    In terms of timing, I agree that it’s a bit late to the game. But if executed well, I think there is still a very good chance. Besides what we discussed, there are still other innovation ideas. I think the end goal is to lift it to be a cross platform system provisioning and management tool.


  3. Alexey Savva
    May 21st, 2014 at 13:24 | #3

    It is already there. I use VMware guest API to configure hostname, network settings, add host to AD domain (Win only) and start Puppet agent. You can upload and download files directly to the virtual machine as well as manage filesystem and running custom programs.
    I don’t think that VMware needs to reinvent wheel and compete with Puppet/Chef/whatever.

  4. May 22nd, 2014 at 11:39 | #4

    Hi Alexey,

    I think you have a good point. There is no need to re-invent a wheel. But nice integration would be good. Also make it work even with no network configured.


  1. No trackbacks yet.