This blog series about WVD has been a long time coming.
When I first learned about Windows Virtual Desktop several years ago, Microsoft was calling it RDMi (Remote Desktop Modern Infrastructure) and we fellow MVPs were all sworn to secrecy about Microsoft’s roadmap for “the new RDS in Azure.” Many of us were excited; it seemed Microsoft had learned from the mistakes they made when creating Azure RemoteApp.
As everyone learned later in the private and public previews, the WVD technology stack was kickass! Microsoft had eliminated the need for organizations to stand up and manage their own servers to handle the RDS infra roles of RDWeb, the Connection Broker, the Gateway, etc. This had been a major pain point for organizations who wanted to stand up full RDS deployments – what was manageable in Server 2008 took on a life of its own with the release of Server 2012, with its completely unintuitive (to this day) Remote Desktop Services Manager in Windows Server Manager.
There would be no more need to buy and install certificates, to fight with SQL Server if you wanted to setup a High Availability RDS rollout, to set RAP and CAP policies on the RD Gateway, etc., etc. They were giving IT departments a big “easy button” to roll out RDS. We MVPs were pumped – Microsoft was listening to its customers and the market! This was going to advance End User Computing. It was going to be great for our industry. Woot!
I was really impressed with the engineering thought put into “RDMi in Azure” and, to this day, I have tremendous respect for the engineers and product managers at Microsoft that brought WVD to life.
“There Ain’t Nothing In This World For Free”
“Oh, there ain’t no rest for the wickedAin’t No Rest for the Wicked by Cage the Elephant
Money don’t grow on trees
I got bills to pay
I got mouths to feed
There ain’t nothing in this world for free.”
I was sold. I started dabbling with the WVD Private Preview, commenced integrating our products with Windows Virtual Desktop and later, when asked, agreed to become one of the original dozen or so founding Windows Virtual Desktop partners.
Now, the following is speculation… I’m obviously not privy to upper-level VP meetings at Microsoft, where they set revenue targets and flesh out licensing strategy. However, I do feel it is well-informed speculation – based on my observation of Microsoft’s public actions in the marketplace since releasing WVD into General Availability.
At some point I suspect a group of VPs got together and pondered how they could leverage this kickass technology to:
- Protect and expand core Windows and Office revenue, through Windows and O365 licensing programs;
- Draw more people into Azure, while simultaneously making it unattractive (or onerous) to leave Azure (e.g. achieve Azure “lock in”);
- Find ways to upsell the new WVD adoptees on related ancillary services in Azure, taking market share from away from traditional partners in the process; and
- Grab market share away from MSPs, hosters, and other colocation firms, by luring their RDS customers away from them into Azure, and capturing that compute and IaaS revenue for themselves.
Microsoft’s stock isn’t doing a moonshot right now because they are dumb. Quite the contrary – they are being smart and pulling ALL the levers they have in the market to grow their cloud business. You have to begrudgingly admire their cunning, it has shades of Bill Gates’ ruthlessness during the Browser Wars of yesteryear.
Look Before You Leap… Unless You Love Big Azure Bills
Before you jump headfirst into WVD, you need to know what you’re getting yourself into and you need to know the real liabilities of WVD. These offset the feel good, sunshine pumping market-puffery that many “technical evangelists” – including some of my MVP colleagues – are churning out at the moment. Otherwise, you’ll be a sheep lining up to get shorn.
I should mention- we at RDPSoft have been adapting our own solutions and continue to adapt our solutions to Windows Virtual Desktop because we know it’s not going away. However, I want to provide some sorely lacking balance to the WVD hype train, before people spend way more than they potentially need to for EUC infrastructure.
You also need to understand and appreciate that running classic RDS on Server 2019 remains a VERY attractive and less expensive alternative for MANY organizations when contrasted with WVD. I’m going to start writing about this more in future posts.
I alluded to some of these issues with WVD when I wrote my “How Not To Lose Your A$$ When Deploying Windows Virtual Desktop” blog, in the Summer of 2019. However, there is much more that remains to be hashed out, which I will do in this series.
To that end, here are some of the biggest WVD gotchas that I see at the moment. I will expand on each of these in depth and add more in subsequent articles.
Microsoft Won’t Let You Run Their Windows 10 Multisession OS Outside of Azure
Nope. Can’t do it. If you want to virtualize Windows 10 outside of Azure, you must do it the old, painful, and expensive way, via a “single user per Windows 10 VM” running on a Remote Desktop Virtualization Host (RDVH) complete with VDA licensing or equivalent. Microsoft won’t let MSPs and hosting partners run it in their datacenters either, and I don’t think that will ever change.
Microsoft Won’t Let You Run WVD Hosts On Your Hardware Or In Your Datacenter, Even Though They Could
My friend Claudio Rodrigues, the crotchety godfather of EUC (You the man, Claudio – you’ve forgotten more about RDS than I will ever know), caused quite a stir on Twitter when he proved, via some creative hackery, that you could route WVD session request through the WVD infra components in Azure on to your own Remote Desktop Session Hosts in a different datacenter, even AWS!
So, Occam’s razor would indicate that the primary reason Microsoft doesn’t allow this is to force you to consume more Azure VMs and related resources, acting as a force multiplier for their Azure billings.