Hyper-V Server 2022

Elden_ChristensenI'd like to add my voice to this thread, and make clear that the precise problem here - for me at least, and likely for most here - is not necessarily a change in product; rather, the problem is the change in pricing.  Microsoft gives away many things for free (email, resources, Azure free-tier products, etc.) and does so for specific purposes: namely, to attract users to the brand and the platform, in the hope that this boosts revenue in the long term.  It's a good strategy; and by removing HyperV Server from its free offerings, Microsoft is potentially losing a lot of "advertising opportunities" of this type in the future.

Most of the people in this thread seem to be people who work at or for big data centers, and provide services to small/medium business clients.  For them, the loss of a free HyperV server has directly geometric revenue effects, but, I'd guess that, at the end of the day, Microsoft believes that these large datacenter people can afford to pay, and Microsoft therefore is choosing to stop offering a free HyperV server because Microsoft doesn't think it's really necessary.  If these companies, Microsoft might reason, want to stay on HyperV, they'll move to the Azure product and pay the fee and get used to it.  That may be true in some cases, although certainly not all.

Now consider the opposite situation.  I am an IT Director overseeing a very small team of a very few people running our own datacenter pair... so I myself am an SMB.  However... I have five *very large* clients.  Let's assume for the sake of discussion that these are huge clients, globally known, who pay my company to operate their IT infrastructure for them, simply because, by virtue of their own structures, it makes more sense for them to outsource such things than to operate them in-house.  In at least one case, they have no house.  My point is, I may be the opposite of what others in this thread might be.

I "grew up" on Linux - I started with SCO Xenix (flashbacks anyone?) back when the i286 was a thing, and moved forward through time to OpenSuse.  When virtualization became a thing, I was on XenLite, and it was... okay... but as we all know it's never been great.  I've always run Windows on my PC, and loved it, but, like many, I wasn't really focused on understanding everything Windows can do.... I was busy running servers, and there were just not enough hours in the day.

Two years ago I purchased a new server for my test lab, and found that it was "designed for Windows".  I was able of course to load OpenSuse on it, and "knew" that I would never actually load a Windows Server on it, because I don't have $10K for a Datacenter license just lying around.   But I also still felt that "XenLite" wasn't really "good enough"... and I wanted to move to something more powerful.  I started sniffing around at XCP-NG, and even had deployed a few Windows Server trial instances on it, when I accidentally discovered Windows HyperV Server 2019.

For the Jeff Foxworthy fans among us, "If a VERSION of WINDOWS SERVER CHANGED... YOUR... LIFE... you might be a redneck..."  

I loaded HyperV Server 2019 on my test lab machine... and never looked back.  HyperV Server led me to play with Windows Servers, in a real way, for the first time - to load, yes, Linux servers as guests, but also to spend tons of time with Windows servers that I'd never otherwise have spent.  HyperV Server led me to participate more closely with Microsoft products, opportunities, and services.  It led me to Azure, and to learn about and become comfortable with Hybrid setups, all the various cloud services there, and more. 

Most of all:  It led me to a place where I was ready to start recommending Windows servers and services to my clients: Linux users who, initially, would have wanted to be on HyperV Server as the hypervisor, but, soon thereafter, would have moved to paid deployments of Windows Server, with its many features and offerings, and, thereafter, I'm sure, to Azure... if I could only have provided them with a reliable pathway so to do... that was within my non-existent SMB budget... and covered by Microsoft's active participation and support.

In other words, there could have been a butterfly effect here.  The others in this thread have made clear that they could have continued to grow their client bases (and therefore Microsoft's revenue) more freely with a HyperV Server 2022 (and beyond!) deployment. In my case, the butterfly effect could have been much more pronounced:  just my own little ability to recommend and influence direction for my few large clients might well have caused a huge shift for them and others to Azure (by way of HyperV 2022, 2025, and *then* Hybrid when the time was right), which could have resulted not only in revenue but also (for the sake of discussion) enhanced visibility for Microsoft.

In other words, continuing to offer HyperV Server 202X could have brought lots of continuing benefit to Microsoft; similarly, this small decision Microsoft seems to have made will bring Microsoft no benefit (at least not that we can discern), and could harm it in terms of significant opportunity costs.

The absence of a HyperV Server 202X, means that that "stepping stone" (or, in my case, "gateway drug" 🙂 is now gone.  Now the distance to Windows Server, Hybrid, and Azure, is much farther away.  Now I'm forced to recommend alternatives, which for me means XCP-NG and XOA because, next to HyperV 2019, that's the best thing out there that fits the profile we're all looking for.  And when you run a Hypervisor based on (*cough*) Linux, there's really no "draw" to Windows or Microsoft from that direction. 

I cannot describe my joy when, at last, I was able to seamlessly integrate my HyperV server, with its running guests, with my Windows 10 ProW PC, and Windows Admin Center, and HyperV Manager, and have everything... just... work.   It was one of the coolest, most reassuring experiences I've had in a long time.

Likewise, I cannot describe my sadness at learning that HyperV 202X may not ever be a thing.

I cannot imagine that it's *difficult* for Microsoft to maintain such a product, it's just Windows Core with some (or many) things limited.  I imagine that maintaining that "version" of Windows Server 202X as jut one additional SKU is comparatively easy.  Certainly easier, I'd say, than maintaining Outlook.com.... or Skype.... or Azure free tier... and so on and so on.   And just as Microsoft gets significant indirect benefit from those things (I for example use Enterprise E5 now, and my gateway to that was Outlook.com, followed by Exchange Server, running on... yup.... HyperV Server), I think it's clear that Microsoft is getting significant indirect benefit from HyperV Server 2019, and that such benefits would continue and even increase with HyperV Server 202X.

I respectfully suggest that it is to Microsoft's benefit to continue to offer that product, and I request - if you have not done so already - that you share these comments and others in this thread with the (other) decision makers out there.  We clearly all think that this should be a thing, and now is the time to make that happen.

Thanks for reading,

Glen