Is the Time for VMware in the SMB Over?


  • Service Provider

    Over the past several years, nearly all recommendations for the SMB (Small and Medium Business) market have been that VMware ESXi / VMware vSphere is the best option for virtualization. The cost, less than $600 for an Essential license covering three typical hosts, was minimal and the product was fast, stable and easy to use. With that cost the product came with nearly everything that a small business could want. Competing products, like HyperV and XenServer, lacked some of the features and the $600 price tag was small enough to make VMware the clear choice.

    The tables appear to have turned, however. With recent releases of both HyperV and XenServer both offer the speed and stability along with ease of use that SMBs need most. Both offer their complete feature stacks for free making them not just $600 cheaper than VMware's offering, but far more featureful for free. Both have increased ease of use, as well, to be on par with VMware's offerings.

    Free means a lot here, that cannot be overlooked. Often VMware advocates point to the fact that if $600 (which can be discounted too) is a troublesome sum for any company, that company has other issues financially and is not viable. This is a valid bit of reasoning. But free carries far more value than simply eliminating the $600 price tag. Being free means that we are not limited to just three hosts or just the features included in the Essentials pack. Features that costs thousands with VMware vSphere are also available for free. Features we would either pay an arm and a leg for or would just do without. Maybe features we don't need today but find would be good to have tomorrow. And being able to add that fourth host is a big deal - this means greater flexibility to simply do what is best rather than trying to work within unnecessary constraints making us more agile and responsive to business needs.

    Then there are soft costs. Acquiring and managing a paid for, licensed product takes more work. We have to acquire and manage licenses, deal with purchasing and procurement and audit for license violations and worry about external audits. We have to do more complicated budgeting. These costs, which are primarily in human effort, add up and are recurring.

    There was a time when VMware ESXi's maturity and technological lead meant that the additional constraints, costs and effort to utilize it were worthwhile in nearly all cases. But it would appear that that era is over. Even at equal costs (if vSphere Essentials was to be made free) VMware's offering pales in comparison to HyperV in effectively all cases and to XenServer in all except the availability of a robust backup API leveraged by free or low cost industry products (Unitrends, Veeam, etc.) VMware's currently free ESXi product is overly anemic to the point that XenServer is always superior and HyperV normally is (unless the goal is to test ESXi itself.)

    While VMware's vShere is an excellent product with amazing engineering behind it, the pricing and licensing model has left it in a position where only the incredible scale of the largest enterprises would justify it as a consideration and even there its value is rapidly waning as cloud technologies make the VMware advantages largely irrelevant even in that venue.

    Is there really a non-niche use case left for VMware ESXi and VMware vSphere in a new deployment in an SMB today? It appears that the era of VMware dominance in this arena has passed.


  • Service Provider

    Beyond the product itself, the ecosystems around VMware are no longer strong than the competition and what used to be VMware's advantage through third party products is now a weakness, at least in the SMB space.

    The biggest products that affect the SMB market around virtualization are backup and replication. On the small scale of SMBs, products like Unitrends and Veeam offer free and low cost backup products that leverage the backup APIs available from the platforms. Because HyperV includes these APIs for free, this often makes the backup free as well. XenServer lacks this same maturity of API and while backups could be done for free, vendors have not jumped on this opportunity yet.

    The other big product is replication and vendors, like StarWind, have free products on both VMware ESXi and HyperV but, in that particular case, with better integration and performance on HyperV. XenServer has this type of replication baked in as well via DRBD. This leaves VMware as the back runner again.

    That the third party or baked in products are becoming not only on par but superior on HyperV and sometimes XenServer is making VMware's traditional advantage a weakness that they are not likely to recover from.


  • Service Provider

    As a side note, in the latest Unitrends release the free version of VMware ESXi not only lacks the backup API so that it cannot be used to backup for free, but the platform lacks features that keep even the Unitrends VM from being deployed to ESXi Free! This means that VMware now has a compatibility issue between licensed versions making it additionally weak and risky compared to its competition.



  • I am nowhere near an expert in this areas but before coming to a SMB, I worked in a huge enterprise with a petabyte+ of SAN storage and we had ESX VM's all over the place...there were some redundancy issues when it came to some SAN errors in the firmware but overall, ESX was pretty solid.

    Here at my one man IT shop and thanks to ML's, I am set up on Hyper-V. :)



  • I've used ESXi and Hyper-V in production environments, both now and in the past, I use XenServer for my home lab. In all honesty I haven't found a major difference in any of the three. I will say that XenServer has been far more stable then ESXi or Hyper-V, although that is in a non-production environment.


  • Service Provider

    @coliver said:

    I've used ESXi and Hyper-V in production environments, both now and in the past, I use XenServer for my home lab. In all honesty I haven't found a major difference in any of the three. I will say that XenServer has been far more stable then ESXi or Hyper-V, although that is in a non-production environment.

    XenServer, being really just an interface and installer for Xen, has Xen to work with as a base and Xen is basically on par far as maturity with VMware vSphere but is used not be medium sized businesses that struggle to run something else, but by the world's most advanced environments (Amazon, IBM, Rackspace, etc.) so the battle testing, enterprise readiness and engineering support is likely second to none and it has had nearly as long as anyone to put that into action. VMware has to do the same work but without the benefit of those massive cloud environments and vendors to back them, and do it all on their own without the benefits of open source. HyperV and KVM are younger which presents its own challenges.

    It's really not surprising that Xen is crazy reliable given that the most demanding environments in the world have been choosing it since the beginning.


  • Service Provider

    @garak0410 said:

    I am nowhere near an expert in this areas but before coming to a SMB, I worked in a huge enterprise with a petabyte+ of SAN storage and we had ESX VM's all over the place...there were some redundancy issues when it came to some SAN errors in the firmware but overall, ESX was pretty solid.

    It's very solid. It is a great product with great engineering behind it. But being closed source and very expensive and lacking any differentiating features leaves it in a position of "when would we choose this" in the current environment of mature, feature-rich competition that is free?



  • You do need SCCM for some of the Hyper-V features to compare with the paid versions of Vshpere.

    The job I've currently been interviewing for is all Vsphere with multiple 42TB SANs. About 10 hosts at the main site. And more than enough servers to failover to with at the 16 other sites, along with Dell app assure replicating between multiple sites. They do have some SANs tied in to azure as well and replicating there. About 8,000 employees in total (and this subsidiary, parent company has a lot more) so Id still call it a SMB.


  • Service Provider

    @thecreativeone91 said:

    About 8,000 employees in total (and this subsidiary, parent company has a lot more) so Id still call it a SMB.

    SMB typically is considered to cap around 500 employees. SME at around 2,000. 8,000 is normally considered a "large" business but not quite enterprise which, I believe, is normally starting around 10,000.


  • Service Provider

    As a subsidiary of that size, it sounds like a "large" subsidiary of an enterprise.

    I've worked for 350 person divisions of a Fortune 10. But at the end of the day, we were still an enterprise.



  • The real question here is will VMware realize this, and give away more for free, or will they stick to their guns and try to convince people there product is better?



  • @Aaron-Studer said:

    The real question here is will VMware realize this, and give away more for free, or will they stick to their guns and try to convince people there product is better?

    I doubt they have to change much of anything at this point, in the future they will try to rebuild their licensing to reflect this new competition. There are so many VMWare users out there that wouldn't switch to anything else. I've talked to several of them who said I was stupid for deploying Hyper-V when ESXi is so inexpensive.


  • Service Provider

    @Aaron-Studer said:

    The real question here is will VMware realize this, and give away more for free, or will they stick to their guns and try to convince people there product is better?

    It's a good question but VMware is caught in a difficult position - VMware sells a product that is otherwise free in the rest of the industry. This means that they cannot give it away or their entire revenue base is literally gone overnight and they have no way to make money. Microsoft, Xen and KVM all have good reasons to be free and have no need to make money from virtualization. VMware is the opposite.

    This is the same problem that caught Windows in the face of Linux. Linux, being free and open source, was able to take over the server market in just ten years and while Windows still plays a very large role, the dominance of a paid option evaporated very, very quickly and holding onto the piece of the pie that they have is based primarily on having a ton of very unique options built into the product and relying on the perception that their competition is complex. VMware lacks both of those things.

    VMware's only real hope, that I can see, is to figure out how to make their product so much better that they are worth the money. Otherwise, they are pretty much stuck. They can't be free and they can't really be easier to use than their competition.


  • Service Provider

    @coliver said:

    I doubt they have to change much of anything at this point, in the future they will try to rebuild their licensing to reflect this new competition. There are so many VMWare users out there that wouldn't switch to anything else. I've talked to several of them who said I was stupid for deploying Hyper-V when ESXi is so inexpensive.

    Much like the Windows versus Linux crowd. Linux can provide nearly any service that Windows can, but the Windows deployment density is heavily (but not entirely) based on having a huge user base that refuses to learn something new or is afraid of change or needs to play politics rather than being financially advantageous to the company.

    Windows, like vSphere, is an excellent product. But the cost for it and the other caveats (licensing overhead, audit risks, extra manual labour, deeper knowledge needed, legal concerns) make it very, very hard to justify.



  • @scottalanmiller said:

    @Aaron-Studer said:

    The real question here is will VMware realize this, and give away more for free, or will they stick to their guns and try to convince people there product is better?

    It's a good question but VMware is caught in a difficult position - VMware sells a product that is otherwise free in the rest of the industry. This means that they cannot give it away or their entire revenue base is literally gone overnight and they have no way to make money. Microsoft, Xen and KVM all have good reasons to be free and have no need to make money from virtualization. VMware is the opposite.

    This is the same problem that caught Windows in the face of Linux. Linux, being free and open source, was able to take over the server market in just ten years and while Windows still plays a very large role, the dominance of a paid option evaporated very, very quickly and holding onto the piece of the pie that they have is based primarily on having a ton of very unique options built into the product and relying on the perception that their competition is complex. VMware lacks both of those things.

    VMware's only real hope, that I can see, is to figure out how to make their product so much better that they are worth the money. Otherwise, they are pretty much stuck. They can't be free and they can't really be easier to use than their competition.

    VMware could easily go the free with paid support option... there are no companies that are doing support for free at this point in time... although they would lose out on less then half their revenue stream.



  • @Aaron-Studer said:

    The real question here is will VMware realize this, and give away more for free, or will they stick to their guns and try to convince people there product is better?

    Add more features or even lower pricing is likely. Free doubtful. They are suppose to be adding things like more virtual networking, in built firewalls and virtual routers etc.



  • Windows, unlike VMware, has the advantage of a massive third party ecosystem that is unique and Microsoft is able to derive value from many different sources like their desktops, MS Office, Office 365, Azure, etc. Microsoft is well aware that the Windows Server value proposition is getting to be very lean and is rapidly diversifying and focusing on higher level platforms to mitigate this risk. This is the nature of the closed source, commodity platform beast and they know it well. Windows Server served them well for a long time but it cannot last forever and will not.

    In time, Windows Server cost will drop until it is not a revenue stream and, hopefully they will wisely go free and license free at that time so that it remains an incredibly bit of technology and will regain much of its lost market share. It might even go open source at that point, but that is a difficult thing to do with code that was never intended to be opened.



  • @scottalanmiller said:

    @coliver said:

    I doubt they have to change much of anything at this point, in the future they will try to rebuild their licensing to reflect this new competition. There are so many VMWare users out there that wouldn't switch to anything else. I've talked to several of them who said I was stupid for deploying Hyper-V when ESXi is so inexpensive.

    Much like the Windows versus Linux crowd. Linux can provide nearly any service that Windows can, but the Windows deployment density is heavily (but not entirely) based on having a huge user base that refuses to learn something new or is afraid of change or needs to play politics rather than being financially advantageous to the company.

    Windows, like vSphere, is an excellent product. But the cost for it and the other caveats (licensing overhead, audit risks, extra manual labour, deeper knowledge needed, legal concerns) make it very, very hard to justify.

    No argument there, the Windows proposition seems to be getting more and more hard to justify with "cloud" models readily available.



  • @coliver said:

    VMware could easily go the free with paid support option... there are no companies that are doing support for free at this point in time... although they would lose out on less then half their revenue stream.

    That's true, but seems unlikely. The revenue drop is probably more than they could withstand. Citrix is already doing this model, as is Microsoft. VMware lacks the additional revenue to make this work, I think.



  • @mlnews said:

    @coliver said:

    VMware could easily go the free with paid support option... there are no companies that are doing support for free at this point in time... although they would lose out on less then half their revenue stream.

    That's true, but seems unlikely. The revenue drop is probably more than they could withstand. Citrix is already doing this model, as is Microsoft. VMware lacks the additional revenue to make this work, I think.

    You think a crucial part of their revenue comes from new installs? I would assume that is minuscule compared to their on-going support/licensing. They are also spreading into attached markets with their VDI manager/infrastructure, Horizon.



  • @coliver said:

    You think a crucial part of their revenue comes from new installs? I would assume that is minuscule compared to their on-going support/licensing. They are also spreading into attached markets with their VDI manager/infrastructure, Horizon.

    Ongoing licensing is part of that same revenue. Both new installs and ongoing licensing would evaporate together.

    For Essentials, support is not even included as it is. The "paid support" model is already there and they still are charging for the initial install as well as ongoing licensing.



  • @coliver said:

    @mlnews said:

    @coliver said:

    VMware could easily go the free with paid support option... there are no companies that are doing support for free at this point in time... although they would lose out on less then half their revenue stream.

    That's true, but seems unlikely. The revenue drop is probably more than they could withstand. Citrix is already doing this model, as is Microsoft. VMware lacks the additional revenue to make this work, I think.

    You think a crucial part of their revenue comes from new installs? I would assume that is minuscule compared to their on-going support/licensing. They are also spreading into attached markets with their VDI manager/infrastructure, Horizon.

    There VDI stuff isn't used much. XenDesktop is a way better VDI solution than VMware's. Also when has anyone needed to use support? Seems pretty rare. It's about like calling Microsoft support. Never need it. I've heard stories online of some people neededing it but don't know anyone who has.


  • Service Provider

    @coliver said:

    They are also spreading into attached markets with their VDI manager/infrastructure, Horizon.

    How valuable is that likely to remain if businesses are forced to "do VDI with another vendor", "do VDI with VMware and everything else with someone else" or "have a uniform environment?"

    I think that VDI and Horizon will do little for them, long term, because the value to that rapidly erodes in the light of everything else. And SMBs do very little VDI and by the time that they do, VMware will already not exist in their market.


  • Service Provider

    @thecreativeone91 said:

    There VDI stuff isn't used much. XenDesktop is a way better VDI solution than VMware's. Also when has anyone needed to use support? Seems pretty rare. It's about like calling Microsoft support. Never need it. I've heard stories online of some people neededing it but don't know anyone who has.

    That pretty much sums it up for me. There is better VDI available from the free players which allows you to have a lower cost, lower risk, uniform virtualization environment on top of the alternative VDI.

    And support I totally agree. If you have an MSP partner, it is they who would use support and not the customer and they have heavy interest in being competent rather than spending money on support whenever possible. Internal IT running one of these products should not need support and the community support is very good if needed. These are super simple products. The only places I see spending money on support are huge enterprises with deep pockets and only do so because of a combination of playing politics (having someone else to blame is better than doing the right thing for the business) or hiding the incompetence of the department (spending a fortune on "support" to hide the fact that the vendor is doing the work instead of the IT guys.)



  • @thecreativeone91 said:

    @coliver said:

    @mlnews said:

    @coliver said:

    VMware could easily go the free with paid support option... there are no companies that are doing support for free at this point in time... although they would lose out on less then half their revenue stream.

    That's true, but seems unlikely. The revenue drop is probably more than they could withstand. Citrix is already doing this model, as is Microsoft. VMware lacks the additional revenue to make this work, I think.

    You think a crucial part of their revenue comes from new installs? I would assume that is minuscule compared to their on-going support/licensing. They are also spreading into attached markets with their VDI manager/infrastructure, Horizon.

    There VDI stuff isn't used much. XenDesktop is a way better VDI solution than VMware's. Also when has anyone needed to use support? Seems pretty rare. It's about like calling Microsoft support. Never need it. I've heard stories online of some people neededing it but don't know anyone who has.

    I haven't had the opportunity to play with VDI much... although I applied for a job that works with Horizon. I can understand where XenDesktop comes into play though it is a very mature software from what I have seen.

    Good point on the support, I guess I am looking at a hypervisor as a fragile piece of software when all my experience points to the exact opposite.


  • Service Provider

    @coliver said:

    Good point on the support, I guess I am looking at a hypervisor as a fragile piece of software when all my experience points to the exact opposite.

    In theory a hypervisor is tiny, does very little and insanely stable. If it is anything else, it should be avoided. All four big boys are great this aspect. This is partially why the Linux Foundation and Microsoft make the hypervisors free.... they do very little and need very little care and feeding. It's a place where if you don't make it free, someone else will (and has.) That there are two, enterprise, open source and free alternatives (Xen and KVM) already shows this. And in the Type 2 space, VirtualBox is free leaving effectively no room for alternatives there either.

    Operating Systems eventually migrate to open source and free over time. Hypervisors do the same but much, much faster.



  • Is the time for on-premise servers in the SMB nearly over? In which case, choice of hypervisor becomes a moot point, right?

    So for me, in the short to medium term, I've invested a lot of time and effort into VMware, so I won't be switching to anything else, and in the medium to long term I'll be running VMs in the cloud, so don't really care about hypervisor technology any more.

    I know cost isn't a factor as $600 is trivial. I've invested far more than that in my time and effort.


  • Service Provider

    @Carnival-Boy said:

    Is the time for on-premise servers in the SMB nearly over? In which case, choice of hypervisor becomes a moot point, right?

    As much as I love moving away from on-premises servers, I don't believe that the era is nearly over. Here is how I feel, without putting a ton of thought into this question:

    • While most workloads in the SMB should not be on-premises, about 10% should be and will remain this way for a long time, slowly lowering but never completely going away. Workloads that will long remain will be heavily cache and security based (proxy servers, DNS, AD, scanning, filtering, machine controllers, etc.)
    • While most workloads should not be on-premises, many will remain because SMB folk have slow upgrade cycles and tend to wear tin foil hats (both IT and managers outside of IT) and do not apply logic and business acumen to these decisions in many cases. So on-premises workloads will exist where they should not for a very long time.
    • Even when moving off-premises to colo, the needs for virtualization choices remain exactly as they do on-premises and colo represents a large percentage of the off-premises server workloads in the SMB and will continue to do so, while slowly increasing as on-premises lowers but slowly decreasing as the move to VPS and cloud IaaS happens.
    • Even in the cloud the choice matters to some degree. Using Xen, while an "under the hood component" allows the best clouds like Amazon to outperform their competition and keeping the pricing low while providing unique features like Xen PV. Using the wrong virtualization platform, while not a problem in and of itself, was a key early indicator that something was wrong with CloudatCost, for example. We knew because they were using VMware that their costs were higher than they should be and their features fewer - which in turn indicated a lack of necessary support skills internally. It is only an indicator, but one that played out very realistically.

  • Service Provider

    @Carnival-Boy said:

    I know cost isn't a factor as $600 is trivial. I've invested far more than that in my time and effort.

    We get $520 by shopping around. And there is huge time and effort in migrating. But you have to include the time and cost of license management, ongoing licensing and loss of features too. For us VMware costs many times the license cost.



  • @scottalanmiller said:

    As much as I love moving away from on-premises servers, I don't believe that the era is nearly over. Here is how I feel, without putting a ton of thought into this question:

    • While most workloads in the SMB should not be on-premises, about 10% should be and will remain this way for a long time, slowly lowering but never completely going away. Workloads that will long remain will be heavily cache and security based (proxy servers, DNS, AD, scanning, filtering, machine controllers, etc.)
    • While most workloads should not be on-premises, many will remain because SMB folk have slow upgrade cycles and tend to wear tin foil hats (both IT and managers outside of IT) and do not apply logic and business acumen to these decisions in many cases. So on-premises workloads will exist where they should not for a very long time.

    Aren't colos often much more expensive for SMB than onsite? Granted, the Colos often provide much better services, the SMBs have been getting away without those features in their locally hosted solutions for 2+ decades now, why the sudden need to change?
    Additionally, the network performance of onsite at 100 or 1000 Mb is considerable compared to running everything from the cloud (colo).

    I want to accept the desire to move to the cloud/colo but I can't see how it doesn't drastically increase costs, and possibly drastically affect performance (bandwidth).



Looks like your connection to MangoLassi was lost, please wait while we try to reconnect.