Hyper-V is better tech than ESXi – What are Microsoft Smoking??

“We are 4x cheaper with better technology versus VMware.”

I’ve been fairly open in my opinion against the latest round of Microsoft FUD coming out of their Worldwide Partner Conference this week but I felt strongly enough by the utter crap coming out of their mouths to respond in a post.

It’s not so much about the claim to be 4x cheaper than the VMware Cloud Suite…but more the outright incorrect claims that their technology is somehow superior to that of VMware’s.

I’ve found myself in the position to have been exposed to both Hyper-V and ESXi (not counting the Management and Orchestration suites) and in fact I cut my teeth in the Virtualization world on Hyper-V…so unlike others out there who see things only through the rose colored glasses Microsoft seem to sew onto peoples faces… I go by a real world operational perspective that’s not blinkered.

So here it is…Microsoft Hyper-V is not the equal or superior to VMware’s ESXi! And rather than go through feature by feature..In the interest of keeping this post short and to the point, I would challange anybody to sit someone who has had zero exposure to the Virtualization market to evaluate both Hyper-V and ESXi side by side…without bias or without prejudice there is no doubt in anyones mind that no logical person would choose Hyper-V as the better hypervisor platform over ESXi. To reinforce that…ESXi will come out ontop.

It’s that simple!

Of course I now fall firmly with the side of VMware and some will argue that my own view is blurred but I can tell you that my current opinions are based on fact and experience…not desperate attempts to discredit otherwise far far superior technology…but then again…Microsoft have made a habit of this so it doesn’t surprise me.

Kevin Turner you are a disgrace!

Read more: http://www.crn.com.au/News/389695,vmware-google-apple-catch-a-spray-in-turners-keynote.aspx?utm_source=feed&utm_medium=rss&utm_campaign=CRN+All+Articles+feed#ixzz37Vca3hea

12 comments

  • You are 100% correct, when you take someone that has no virtualisation experience I think that VMware is better than Hyper-V, not by a lot though.

    Virtualisation is so 2005 though, when you start talking about Private, Hybrid and Pubic cloud (HyperVisor + Management Tools) Microsoft is streets ahead in terms of completeness, integration and capabilities. IaaS will pave the way for PaaS in the coming years and in this space VMware dont have a hope in competing with Azure.

  • You’ve got it all wrong…. clearly KVM + OpenStack is far superior on all counts…. especially on maturity in the enterprise #troll

    But (slightly) more seriously, I find it really funny that only 12-18months ago MS’s playbook for Hyper-V was “Ok, so its not great, but its #GoodEnough (PatentPending) and free… kinda” and now, assuming that vSphere has stood still, they’ve caught up…. Yes it happened w/ Novell, but I don’t see the same scenario playing out here… at the very least, not yet.

    To me, hypervisor / virtualization platform decision always boiled down to a few scenarios:

    * I want a mature, feature-rich, out-of-the-box product that just works and am prepared to pay for it (and 4x is not reality… lets be serious).
    VMware vSphere

    * I don’t really care, free is good, flexibility would be nice, I’ll deal with immaturity (in other words: shift commercial software budget into devops salaries) :
    OpenSource solution (KVM + OpenStack, most likely)
    + Bonus: In charge of own destiny.

    * I don’t really care, free is good, I’ll deal with immaturity:
    Hyper-V

    And PaaS/SaaS…. really? How many webscale properties are based on MS products at all (let alone Hyper-V)…. Hint: MS themselves… that’s about it.

  • I don’t agree Stephen. Server Virtualisation is 2005 but Storage and Network and Network Functions Virtualisation are still maturing and really coming to the forefront. In terms of a Hybrid Cloud, server virtualisation as much as it may be old news, it is still a fundamental part and basis of a cloud platform.

    The fact VMware is the market leader and feature rich in comparison to MS from a Server/Network/Storage Virtualisation aspect, means VMware will have a cloud offering that will compete with Azure. VMware is way ahead of Microsoft in race for the SDDC and simplicity and ease of management goes hand-in-hand when a company can offer all the products and solutions that integrate with ease.

    A large business will more than likely run an internal cloud based on VMware, why would they operate a Hybrid Cloud on two disparate solutions i.e. an internal VMware Cloud and an external Azure Cloud, this is where management and compatibility becomes an issue. Is a workload going to operate once it is move to or from each platform?

    Microsoft’s target market remains SMB, the issue they have is they do not know what their core business is or where to concentrate their efforts. Microsoft tools do integrate well…with their own tools, but the tools are usually complex to set up or they just don’t have the tools to do the job you require. This is why until they address this, they will not compete with VMware.

  • Frank Anderson

    Great response Oli..From the looks Stephen has missed the point of the post and taken a typically ignorant Microsoft only view of the IT world.

    To claim Virtualisation as being so 2005 shows that he really doesn’t understand the industry…Microsoft was late to the game even back in 2005 with Virtual Server 2005, which was a type 2 hypervisior…Virtualisation didn’t really take off and become mainstreamly accepted as a trusted platform for all applications until 2009-2010…there was a good deal of maturity that had to be had in the type 1 hypervisors up to that point…but still Hyper-V lacked in features and reliablity then…as it does now!

    We are moving into an enhanced Virtualisation time where as your correctly pointed out, SDDC will continue to improve along with storage Virtualisation and now network Virtualisation …how’s Microsoft doing in that space?? Similar story…they are behind again! NVGRE is a laughable approach to virtualised abstraction of the physical network layer.

    And to claim that Azure will wipe the floor with VMware’s Cloud also shows ignorance…Azure’s biggest competitor is AWS …a true PaaS provider that covers all aspects of webscale type architecture and design…Microsoft are getting desperate and they have already lost their grip on their traditional strongholds in application and server software…that’s why they resort to gutter tactics and plain outright brainwashing. In fact your comments sounded like they where lifted straight from the Microsoft Coolaid…

    If you like lipstick on pigs…then you will like Azure 🙂

  • Lets see if I can sum up the VMware fanboi arguments here and respond:
    – SDDC is the new shiny. No argument here but I’m yet to see a convincing implementation by anyone other than Google or maybe OCP. Neither VMware nor Microsoft has produced anything satisfactory in anything other than a showcase implementation.
    – Large scale private clouds in production now are likely to be built on ESX. This is probably true today but is missing the point. I’ll come back to this later.
    – Microsoft management tools only integrate well with other MS tools. Even if this was true, and I’d argue that it isn’t, name me a virtualization/cloud/orchestration/whatever product that you can manage cleanly with VMware tools that isn’t another VMware product. And complex to set up? I have 3 letters for you. SSO. Stones. Glass houses.
    – Microsoft’s target market is SMB. First, irrelevant. Second, wrong.
    – Hyper-V used to be bad. No argument. 2008 Hyper-V was awful. 2008 R2 was only marginally better. 2012 Hyper-V was ok. 2012 R2 is competitive and feature comparable for 95% of use cases.
    – Public cloud. I’m really not sure what your argument is, here. Azure isn’t competing with a VMware public cloud because there isn’t a VMware public cloud. You’re right, Azure is competing with AWS, which is running on Xen. So VMware’s dog in this fight is … ?

    Here is the core of my argument about why VMware should be worried about Microsoft eating its lunch. I 100% agree that most significant private clouds today are running on ESX. The very next thing that almost all of those private cloud owners do after paying for ESX is turn around and pay Microsoft for a bunch of DataCenter licences. If Hyper-V is even somewhere close to comparable, one of those steps is unnecessary and it’s not the Microsoft one. If you go to a private cloud owner today and say “I can make it so you don’t need to pay for VMware any more”, you’ll be fending off offers of first born children before you get anything else out.

    Here are some points to consider. Microsoft is not only embedded in the server virtualization space, it functionally IS the platform being virtualized. That’s not going to change in the short or medium term. Hybrid clouds offer a bunch of traditionally expensive services at very low cost (I’m thinking backup/DR/HA) and are significantly complicated by going cross-product. There isn’t a VMware public cloud.

    In my opinion, the only thing that VMware has going for it in the private cloud space is momentum. ESX is in place and changing is difficult (and is made more so by the fact that most people working in virtualization have opinions very much as described by the comments above). Once someone does the sums on exactly how difficult, and how much they would save by not having to pay for VMware, I think VMware should be very very worried.

    • Mike,

      Take the “fanboi” flamebait out of your post & I agree with your arguments. I think the real problem for VMware is not that Hyper-V (or Xen / KVM / whatever) is as good as, or better than, ESXi, but rather that Hyper-V, etc. are *good enough*.

      From a purely “dollars and cents” perspective, once you have *good enough* technology, you don’t gain much by being *better*.

      For the cloud argument … there are basically two “kinds” of cloud: Enterprise & Web Scale (yes, I am ignoring IaaS/PaaS/SaaS/private/public/community/hybrid…). The best exemplars of each, that are recognizable to pretty much everyone with a pulse, are Azure and AWS. The key differentiator between the two kinds of cloud is a very foundational design consideration:

      – Enterprise cloud is designed to primarily target enterprise applications. These are typically stateful applications that depend on a reliable infrastructure. They often do not take kindly to having an “instance” fail underneath them. This is where you’ll find the majority of legacy line of business (LOB) applications that end up being moved to the cloud.

      – Web Scale cloud is designed for applications that are designed and built to be run on an unreliable cloud infrastructure. Don’t believe me? Check out the webinar from 2 May 2013 available here https://aws.amazon.com/apac/webinars/ and see the video at http://youtu.be/X89nZhXJ6mU. The fact that the cloud infrastructure is “unreliable” means that it can be cheap — you’re moving the intelligence up from the infrastructure and into the management and/or application layer.

      When you look at the cloud landscape today, you’ll find that most of the Enterprise cloud offerings are based on *either* VMware or Microsoft technology. Since I work for Virtustream, I’ll use them and Azure as examples of Enterprise cloud. You all know that Azure is backed by Hyper-V. At Virtustream, we use VMware as our primary hypervisor solution; however, we also offer support for KVM via OpenStack, and Virtustream xStream (our cloud management solution) is hypervisor agnostic, so can support a wide range of technologies.

      When you look at Web Scale cloud, you’ll find that most are based on open source technologies (AWS, Google Compute Engine, OpenStack, Eucalyptus, etc.). VMware is noticeably absent from this list. Why, you ask? It’s simple: if the infrastructure you’re building doesn’t need to be resilient (or the resiliency is built into the management / application stack rather than the hypervisor layer), you *need* the hypervisor to be as inexpensive as possible.

      So … what does this mean? Cloud usage is stratified into two camps: 1) those users who want to move legacy applications to a cloud environment, and 2) those users who want to build new applications to exploit the capabilities of a cloud environment. Today, none of the “big three” cloud providers (Amazon, Google, Microsoft) provide a perfect fit for both use cases. Amazon & Google offer good solutions for Web Scale, and Microsoft offers the best solution (of the big three) for Enterpise. Where does that leave VMware? Holding the metaphorical bag. VMware offers the “best” (whatever that means) technology and the most complete feature set, but:

      1) when it comes to Web Scale, VMware is “too good” and thus, too expensive
      2) when it comes to Enterprise, VMware doesn’t have a jockey in the race (they’ve got a horse (ESXi / vCloud suite) that a lot of people are riding, but they don’t really have a complete offering of their own).

      The good news for VMware: it’s much easier to implement Web Scale functionality into an existing Enterprise class solution than it is to implement Enterprise class capabilities into a Web Scale environment.

      The problem for VMware:
      1) Web Scale cloud consumption is growing rapidly. VMware can’t compete in this space.
      2) For Enterprise cloud, Microsoft already has their own horse, jockey, and stable in the race with Azure. Is it technically “as good” as VMware? Probably not. Does it matter? Not really. Public cloud consumers are not buying a technology. They’re buying an SLA.

      The saving grace for VMware is that there are a lot of vendors (including Virtustream) who rely on VMware’s technology to support their Enterprise offerings. Although I will say that many of those vendors are constantly looking for ways to drive cost out of their solution stack.

      So, is Hyper-V better than ESXi? IMHO, it doesn’t matter.

      • Absolutely agree Ken.

        The only thing that I would add is that I see PaaS being much more the future than IaaS. Even down to the work Microsoft is doing to support clustering etc in main-stream servers rather than high-end, this will feed down the stack until the web scale model (compute as a utility, fault tolerant services/applications, zero concern for individual servers) is just how things work even at small scales. Which is even worse news for VMware.

        • The point of my post wasn’t to engage in a discussion over PaaS vs IaaS…keeping it very simple I was commenting on the outrageous comments that Kevin Turner made during his keynote. I used it as an example of the FUD Microsoft creates to fight the fight.

          Comments on Virtualization being soooo 2005 and PaaS being the future over an insignificant IaaS sector are welcomed…shows a high level of ignorance to be honest.

          Myself, I have been able to continue to evolve my career path and follow and embrace Hosting and Cloud technology trends. I’m lucky to not have been exposed to enterprise mentality which is where I see a lot of the Microsoft will be king comments stemming from. In truth I’ve possibly helped pave the way for Microsoft Azure with the work I was involved in as a SaaS Provider doing Multi-tenant Exchange, SharePoint and CRM as well as Web Application and Database Hosting…Traditional Hosting Providers paved the way for Microsoft to launch Azure and make it successful…in the mean time they screwed over their partners…but that’s another thread.

          http://www.microsoft.com/casestudies/Case_Study_Detail.aspx?CaseStudyID=4000003849

          Who’s a fanboy now 🙂

  • Narrow minded to think that MS are the only company that matters. The sands are shifting below the feet of them and they can’t milk their desktop and enterprise clients forever, or continue to be late to the party. They do realize they are 2nd place when it comes to the servers that run the web and this will continue as the shift to large web based services expand at a rapid rate. It makes sense for those companies developing these services to be platform agnostic or develop on their native platform as said above which in many many cases are NOT MS.

  • I can only relate my own operational experience. I work for a company that is, potentially (I have no means to verify, but MS drops hints every once in a while), the largest deployment of Hyper-V clusters in the world. Let me just say that Hyper-V as a technology is nowhere near VMware as a serious competitor. Issues with Hyper-V are consistent, and SCVMM only exasperates the situation. I feel that I can say, definitively, that it cannot be considered as a real choice for enterprise virtualization (that is, unless you are willing to accept all of the caveats of its operation, and nearly constant outages).
    I would gladly take up this point with ANY Microsoft employee/Executive involved with Hyper-V. All of the evangelists who spout off this nonsense have never witnessed it being used in this capacity.

  • Duke Walters

    I agree with this last post about Hyper-V and scalability issues. But let’s face it. At this scale even VMware has problems. I work for a company with 156 ESXi hosts in a live environment. There are many issues with scalability and reliability. We resorted to writing our own tools. To be fair, at this scale you are going to have problems with any hypervisor.