Microsoft Office and the death of RDSH.

Thanks to the announcement at Ignite of Windows Virtual Desktops (WVD) and the information or more details that followed, specifically regarding Microsoft Office on the platform, several people in the industry are now announcing the death of Remote Desktop Services Session Host (RDSH).

Before proceeding with what I think, let me clarify that I do not think RDSH is the silver bullet. It is, like anything else, another tool for delivering applications to your users. That said, what we know as of today is very simple to summarize:

  • RDSH is heavily used to deliver Line of Business (LOB) applications and many, if not the vast majority, have ties to Microsoft Office and/or interact with it.
  • RDSH has been around since 1998 when it was officially announced at the New York Expo. That means we have been dealing with it for over 20 years now. We know how it works, how it behaves and its weaknesses and strengths.
  • As we know it, we learned over the years how to deal with applications that do not like RDSH as a delivery platform. And by we, I mean thousands of people on the internet. You have an issue with RDSH, someone can probably fix it within a day or two.
  • We clearly understand its scalability, performance and for many companies out there, the most important thing, its cost.

Now, looking at WVD, recently announced, all the above is completely unknown. How does it perform at scale? How does it work with all the LOBs out there, that as of today live in a happy RDSH world? And how much does it cost per user per month to deliver the same performance, scalability and availability as all these RDSH environments out there? Yes, we have no answers to that. Yet.

Then Microsoft announced that Windows Server 2019 and its RDS incarnation, does not run Office 365 but does run Office 2019 perpetual. Once people read that, the death sentence for RDSH was issued.

But does it make sense that to be the case? IMHO, no. And a big NO. And actually, quite the opposite.

What we know is simple: Office 365, like its Windows 10 counterpart, is or will become a nightmare with its almost daily release cycles. Yes, the same Office 365 you are running on your PC is not the one that was there last week or a month ago. It may even have new features.

Imagine that with critical LOB apps that interact with Office. If one thing changes on the Microsoft Office stack, everything may break on all these apps that rely on it. Reason why LOB apps are treated in a very strict way regarding patches and upgrades. For most businesses these are critical. If they go down, revenue could be directly affected (like one customer I have where one hour with their main LOB down means USD 1M in revenue lost). They do not like insane release cycles. They do not change their hosting environment on a weekly basis.

The other key point is, what are the exact differences between Office 2019 perpetual and Office 365, other than fast release cycles? What exactly is lost, functionality wise, when I run Office 2019, compared to Office 365? And the key question here, do these features are required by all the LOB applications that interface/interact with Office? Based on my experience, the vast majority of these LOB apps could not care less about whatever collaboration or online features Office 365 brings to the table.

For new deployments, mostly Office apps based, I do and clearly see the reason to go for Office 365. These are new deployments, not having to support mission critical LOB apps. Online features and collaboration are appealing to these use cases. Awesome.

With all that said the reality is, all the LOB apps will not be fixed overnight to work with Office 365. Even if they work, testing and certifying these apps with it may take ages and potentially may never happen for several companies. They will simply stick to what is known to work and to work well. This on itself may provide RDSH the fuel it needs to keep running, and existing, for many years to come.

And if the company all the sudden needs everything Office 365 has to offer, simple. Use WVD to handle that and keep RDSH for the mission critical LOB apps with Office 2019. Note that I did not mention cloud or on-premises anywhere and for one simple reason: RDSH does run anywhere, including the cloud. And unlike WVD, it runs on ANY cloud. Azure, AWS, Google, you name it.

The flexibility anyone expects from a mature solution. And in case you were going to the RDSH funeral, turn back and go home. 

And by the way, 2019 is not the year of VDI.

CR

1,532 total views, 9 views today

RDS – The Complete Guide book. Present and Future.

Ladies and Gentlemen,

As you realized, the book I wrote with Freek Berson (awesome guy, great Microsoft MVP for RDS) is now out on Amazon, in both e-book and paperback versions. It is available here. Before you ask us or mention anything about the book, there are a few things I do want to explain the reason behind. So let’s take a look at some of these:

  • Book sucks as it is about Windows Server 2012 RDS. Well if you noticed, RDS on 2016 is pretty much identical to RDS 2012 R2. Very few differences. Same for Windows Server 2019, now in technical preview. Add to that there is way more production RDS environments based on 2012 R2 than 2016. Finally, the book provides a solid foundation that applies to RDS in general, not tied to a particular release. Learn all that is in there and you will certainly be able to deliver a solid RDS solution regardless of the version in question.
  • Certain chapters were sponsored what means you are biased. Well if you know me well from all the conferences I participated over the years, or as an MVP, you are probably aware of the fact when something sucks, I will be the first one to let you know. Over the years I told Microsoft everything I thought was horrible with RDS. Same with Citrix and its products. Did not matter the fact I was a Microsoft MVP for 17 years or a Citrix CTP for 9. For the book, after trying many products to address the RDS shortcomings (and there are indeed several), we truly believe the ones in the book are some of the best out there. And more than that, these are products we use on a daily basis. That is the only reason we reached out to these sponsors in the first place. It was not just about money to allow us to focus in the book but more like bringing to the public what we think are the top tools out there to make your RDS deployment much better.

With that in mind, so what about RDS 2016 and 2019? We are covering these indeed but as companion books. The idea is all the foundation you need is in the RDS 2012 book. The first one. Then, based on what your target platform is, 2016 or 2019, you just get the companion book to complement all that you learned from the main book. Simple. Also this allows us to release these smaller companion books at a much faster pace than what took us to get the RDS 2012 one done. The companion books to be released are:

  • RDS 2016.
  • RDS 2019.
  • VDI using Microsoft RDS.
  • Deploying GPU-based solutions with RDS.

These four should then cover RDS end-to-end, no matter the version and the scenario you have in mind.

I hope this clarifies a bit about the book and the plans we have for it down the road.

Cheers.

CR

899 total views, no views today

RDS Modern Infrastructure. Modern?

As tons of people spend the week at sunny Orlando for Microsoft Ignite, here I am sitting at home, reading all these tweets and posts about what is next for Microsoft’s Remote Desktop Services stack, RDS for short.

If you read any of these, you are probably aware that Microsoft is changing RDS for the better (hopefully) and the new platform is being called as of today, RDSMi, a pretty term for ‘RDS Modern Infrastructure’.

The more I read about it, the more I think Microsoft has very little clue on what they have been doing with RDS since its early days, dating back all the way to 1997’s Hydra beta availability. And after seeing this ‘RDSMi’ acronym, I can also say with a pretty good degree of accuracy that marketing and its army of marketeers, are deeply infiltrated on anything RDS. As usual, I can certainly and clearly explain the reasoning behind my assumptions.

First of all, if you are not aware of that by now, I have been in the RDS business for quite some time. By that I mean I was probably deploying RDS for customers way before you got a degree and left school. ‘You’ does include many people in the RDS team in Redmond. And being an RDS MVP since 2001, I have seen it all at Microsoft for a very long time (16 years straight, yes, that long). Not only me but others like Benny Tritsch and even Alex ‘Bozo’ Cooper have experienced the same.

So what is the issue and why I am writing about this? Simple.

One of the biggest things the marketeers out there are now promoting and saying about this incredible ‘RDSMi’ thing is the fact many components now do not need to be domain joined. On top of that, if I am not mistaken, there is also an agent of sorts that is now on your RDS Session Hosts.

In other words, RDSMi is basically what we have been telling Microsoft that RDS should be in the past 16 years. Yes, that long. After getting tired of seeing nothing being done, back in 2003 we actually wrote AND released to the market an RDS Gateway that, guess what, was NOT domain joined! Probably sorcery and witchcraft but somehow I managed not to be burnt alive as a witch or warlock. If Microsoft is naming this new thing RDSMi, what was WTSGateway back in 2003? RDSFVi (RDS Futuristic and Visionary Infrastructure)? So please, there is nothing new or modern here.

What is even worse is the simple fact all this shows how Microsoft (and several other vendors in this industry, Citrix included) ask for feedback from MVPs, CTPs and so on and refuse to take it. Taking it 16 years later, at least for me, does not mean you took my feedback. They simply ignore the fact that people like you and me not only have been in this industry for probably way longer than most of the people in these teams but also that we are the ones architecting AND deploying such solutions in the real world. The hands-on people. Very different than saying ‘we listen to our customers and partners’ when what that really means is ‘we pay third party companies to do some research for us and this is what we got from them’. WITHOUT EVER DEPLOYING YOUR SOLUTION IN PRODUCTION, AT SCALE. Funny.

Resuming, and not to ruin your week at Ignite, Microsoft, especially in the RDS space, is just doing what many people told them over a decade ago. Nothing new here. I have to say I am not that easy to impress. But this, seriously? Good try. Maybe on the next Ignite.

For that reason, I am renaming ‘RDSMi’ to ‘RDS Meh Infrastructure’.

And marketeers out there, I am available in case you need some better marketing work.

CR

2,092 total views, no views today

RDS-O-Matic for Windows. It is ready!

It is show time people. Just want to show you the native RDS-O-Matic Win32 app. I had time to port if from the Excel spreadsheet idea and it is not only small but works like a champ.

I know Alex (Twitter @E2EVC) will be complaining about its GUI but hey, for a tool that is this simple there is no need for a fancy GUI. People want apps that work and are simple to use. That is the bottom line and this is what you get with RDS-O-Matic.

So here you have it, on all its glory.

Just want to remind you guys one more time that RDS-O-Matic will be officially released at PubForum (E2EVC) Dublin, in June 2016.

Cheers.

CR

 

14,329 total views, 2 views today

VMware Horizon 6. The only article you will ever need to read.

Ladies and Gentlemen,

We all knew this was going to happen and it happened yesterday. If you have no idea what I am talking about let me quickly summarize it for you and then give you my take on it.

VMware announced yesterday that it is adding support for Microsoft RDS Session Host (a.k.a. Terminal Server, Terminal Services, TS or simply RDS) on its product. So now they can deliver sessions from either Desktop OSs (what VMware View was all about since day one) and from Server OSs (with the RDS Session Host role enabled) using PCoIP.

Why I am saying this is the only article you will ever need about the subject? Well first of all I am the one writing it. Does not get better than that. Then I am not on VMware’s or Citrix’s payroll. Finally I am one of the so called ‘Dinosaurs’ in the RDS world (remember, I got the first MVP award ever for RDS specifically back in 2001). Oh and I drive a Lamborghini.

So seriously let’s take a look at the whole thing and what I think it is important with this release.

– RDS as a platform. I am very happy to see VMware doing this. Honestly. This just proves that all I have been saying all these years, that RDS is a solid platform AND not going to the grave in the near future is true. VMware now officially recognizes this. This also means a lot more work for all of us in this industry as now lots of VMware customers will start deploying this and will realize it is way more complex than a broker and a protocol. They have to deal with printing, profiles, logon times, session sharing, etc. The list goes on. For us, the industry dinosaurs, this is GREAT news. Be prepared to have hundreds of new customers lined up at your door, asking you to help them with their RDS issues.

– Citrix as a solution. There is no other way to put this. VMware is validating what Citrix has been saying for years WHILE acknowledging they (VMware) did have a big hole on their application delivery solution and that Citrix was correct all these years by addressing both the desktop and server OS application delivery mechanisms. Yes, a little tap in the back for Citrix.

– Citrix as a company. One thing I have been saying to Citrix for YEARS, even though I am a Citrix CTP as well, was the fact Citrix was milking the XenApp cow for VERY long, without really innovating much. Minor improvements here and there, evolution (albeit slow IMHO) instead of revolution. Then the world, according Brian Madden, would flip everything to VDI and RDS would die, Citrix jumped into the VDI bandwagon and more than that, started to back stab the product (XenApp) that made Citrix, well, Citrix. Decided to rename XenApp to XenDesktop “Customers are stupid” Edition (ok, App Edition), chop off some features that made XenApp 6.5 a very solid platform and then released XenApp 7.5 “Phoenix” again, still a limping version of XenApp 6.5, not really offering anything better than its previous release. Basically screwing its customers, partners and itself along the way. Cannot get better than this, screwing up wise. Not sure who they hired for the job of screwing things up but whomever that is, this guy is a GENIUS at the subject. Next time I want to screw up something I will definitely give Mr. G a call.
So VMware announcement means two things for Citrix: first, RDS is indeed an important platform what leads to XenApp is important and has to be fixed, if you do not want people starting to test Horizon 6 to jump ship or not buy your product. Secondly, and the most important thing here is, Citrix now has someone on their back and if they want to stay on top they will have to become the Ol’good Citrix we, the dinosaurs in the industry (RickD, DougBrown, SteveG, SBass, Benny, etc) learned to love. The one that innovates, that pushes the industry as a whole forward. And not the current Citrix that looks more like a bunch of farmers that know nothing more than milking a cow. And supervised by a marketing clown. Yep, it is that bad. Hopefully this will be great for the industry, leading to the same type of war we saw at the protocol level, where years ago Citrix was the king by a huge lead and now for 99% of the use cases the protocol is almost irrelevant (this helped the industry so much that even Microsoft released something great, RDP8.1, what is something borderline mystical as they do have a history of releasing stuff from their asses – you know what that is). So the lesson here: this is great for the industry, great for Citrix – if they see this as a challenge and live up to the expectations – and great for VMware, that is broadening its reach and addressing the problem properly. Great.

– XenApp as a product. Well thanks to customer feedback (more like customer wrath really) Citrix had to bring it back from the ashes. Then VMware comes and tells the world RDS is amazing. I hope this is a wake up call to Citrix so they realize how important XenApp is and always has been for their strategy and more than that, for them as a company. This move by VMware hopefully will guarantee XenApp is a product customers can trust in the long run, what many feel was not the case since Citrix almost renamed itself Cindesktop.

– Horizon 6 itself. If you have been in the industry for long you know there is more to RDS than simply having a way for people to connect to an RDS Session Host over a protocol. Problems that are not there with VDI (app compatibility, session sharing, etc) will definitely be there when you throw RDS to the mix. Right now, no one has played with Horizon 6. No one knows what it can do as a complete solution, as something that goes beyond brokering a session to an RDS SH host using PCoIP. How does it handle printing? How does it handle the user environment? How does it handle the server build itself? How much automation there is to increase farm capabilities? The list goes on and for now no one has an answer to that. That is why no decent blogger should say Horizon 6 is great or it sucks. No one knows that. And I can bet things will change from what some analysts saw today to what will be actually shipping. My take is, if VMware is intelligent, they carefully looked at what is out there, the competition, and addressed most of the needs when it is out. If that is not the case, customers may get burnt with a solution that falls short from its promises and may go for a competitor. Or, if you are really loyal to the brand and NOT in a hurry to have that working, you may just say “Oh well it is a V1 product so half of the things not working properly is to be expected – they will get better”. My personal take is I hope it is good as again this will drive the competition and the industry forward. And I will have years of consulting on the RDS space still to go. Great. But until I see it in the wild I cannot say how good or bad it is. Period.

– UX is important. Yes, the user experience is key. And how seamless things integrate with all the platforms that can work as an endpoint is very important. As Shawn Bass mentioned, Citrix ignored a lot of platforms with their receiver, to the point the receiver on OSX for example sucks. I will say this is an industry trend in general as Microsoft apps on OSX do suck too. But there is one point we cannot forget: the AX (the Admin eXperience) has to be good. No matter how good the UX is, if the AX sucks big time and the whole thing is a PITA to get going and to maintain, IT departments will certainly back slash it and bury it somewhere. Lesson here is it has to be polished in all fronts, especially if you are the last player to the game, the one that had years of research available, studying everything that sucks with your competitors. So yes, we do expect VMware offering to be polished in all fronts.

– VDI as a platform. Well thanks to the first point on this article, Horizon 6 puts the last nail in the VDI coffin. What I mean is, in the coffin that says VDI is everything, VDI is better than sex, I want to do a MILF with a VDI tattoo on her lower back (I bet you pictured it). VDI is simply another option, another tool in your tollbox and VMware finally acknowledges it. Plus this goes beyond Citrix and VMware. This is also a wake up call to all the VDI fanboys out there, that were blinded by Brian’s predictions (failed by the way) that VDI was going to take over the world and Claudio would retire due to lack of work for him as an RDS guru. Lesson here, VDI fanboys, go learn RDS and stop thinking BrianMadden.com is the bible. Brian is no Jesus. He does not even have a long beard. And he lives in San Francisco.

To conclude this post I just want to say this: 2014 is the fucking year of RDS and this is not a prediction.

Thanks VMware for confirming what I have been saying all along.

And VMware, welcome to the RDS world. I have my arms wide open.

[Hugging sound]
[VMware fanboys crying in background]

CR

18,121 total views, 2 views today

BriForum Boston 2014

This week Brian and Gabe announced the sessions for both BriForums (London/US). I am happy to announce I will be presenting two sessions in Boston and will almost certainly attend BriForum London as a regular peasant.

If you did not read the list of sessions, here is what I am presenting and why I think these will be useful and what the plan is regarding delivering them.

SBC Round Up 2014. I really like doing these. Plan is to go through the installation and testing of several RDS add-ons (i.e ProPalms TSE, Dell vWorkspace, 2X, etc) and see how the compare to each other and of course to RDS 2012 R2 by itself. What will change this year is I am actually recording all the installations and will post all videos as soon as BriForum Boston is over. Also creating individual PDFs for each product installation so at the end you will get an end-to-end guide on how to install every single major product out there. Neat.

RDS-O-Matic. This is basic the end-result of dealing with RDS installs almost on a daily basis for customers around the globe. The idea was to come up with an automated way to create all the PowerShell commands to deploy a full RDS 2012 R2 from scratch. For BriForum it will be able to perform the following tasks:

– Hyper-V only.¬†Creates all the required VMs based on a sysprep’ed VHD. Of course this requires minimum services to be up and running already like your AD, your Hyper-V hosts, the clustering, etc. But if these are there you simply select the VHD you want and it will copy to all the required VMs, mount them, inject the Unattended.xml file and finalize the setup (add to domain, set IP, add to proper OU). This is optional (meaning if you do have all VMs ready to roll you can opt this step out). Yes before you bitch I have no love for VMware ESXi anymore.
– NLB. For every component that needs NLB you will be able to choose if you want it done for you (i.e. RD Gateway). It will create the VIP, add the ports, etc.
– UPD. If you want to enable the User Profile Disk on the deployment.
– Whole deployment. Of course it does that. Sets up the connection brokers, web access, gateway, session hosts, etc. The whole deal.
– SQL Bullshit. Ideally I will try to automate the turd Microsoft created when setting up the SQL for the Connection Brokers HA. It is a PITA (create folder on the SQL, create database, add proper security, etc – amazing how every other product on the market can do this but NOT Microsoft).

The main plan is to turn all this into a web service that anyone can hit, enter the information and get a text file ready to use for the whole deployment. Later iOS and Android apps so you can do that anywhere/anytime/offline.

And for the first time in 10 years of BriForum for me, I will be actually driving to Boston this time what may actually be faster than flying, assuming the cops do not stop me in Maine. Feel free to stop me and say ‘Hi’ if you see me around at BriForum. I will be driving ‘Ferrucio’ (yes, my kids name all the cars we have at home).

Lamborghini Gallardo

Cheers.

CR

5,630 total views, no views today

Adobe Reader XI on RDS 2012 R2

Just a quick post regarding a stupid issue I had this week with the latest Adobe Reader XI running on RDS 2012 R2. After installing it when trying to launch I would get this error:

Adobe Reader XI Error
Screenshot thanks to screwloose.com.au.

The usual fix is to set the bProtectedMode registry key to 0. Problem is even after doing it the application still refused to launch.

Once I set the Compatibility Mode for the app to ‘Windows XP Service Pack 3’ I was able to launch it successfully. So in case you have the same issue make sure the registry key is there AND the app is set to ‘Windows XP Service Pack 3’ compatibility mode.

CR

3,917 total views, no views today

RDS Projects

As many of you know I am kind of biased towards RDS, especially with 2012. That said it does not mean I will recommend or force down your throat RDS for all use cases. Our main concern as a company is to always provide the correct solution to a problem/business need and never the other way around, seen many times in this market, where they try to find a problem/business need that matches a particular product that has to be sold to the customer.

With that in mind, regardless of the product or solution to be used, it is important you do your homework before attempting to come up with a pilot. That is where things get weird with RDS. As it is now completely part of the OS, what means you do not need to call a reseller or download crap from a vendor, most people think “Well it is part of the OS and I know Windows so I am sure I can pull this RDS thing off” and guess what, these attempts either fail or create a wave of tweets saying RDS stands for “REGRET DAT SHIT”. Not the case.

There are many reasons for that. RDS, even though it is right there on the OS and can be installed with a couple clicks, it is still RDS. That means many things are still at play like RDS Flags, Shims, Profiles, Load Balancing, certificates (ahh these damn certificates) and so on. Exactly the stuff most people are not aware of or simply do not understand.

Add to that tons of resellers out there that need to sell product A or B to keep their Silver/Gold/Unobitanium status, and we have a mess. Now without the knowledge of what is under the hood (RDS) they are now recommending and saying the solution is product A or B and that RDS is shite. BS.

One example (real world) we had was a major company, owners of several high end brands in many spaces (from watches to wineries) that approached us regarding a Citrix project they were working on. Before moving ahead they wanted to understand if Citrix was deemed necessary or not.

After reviewing the whole thing with them I was confident we could pull it off with plain RDS and as XenApp runs on top of RDS I told them to go from pure RDS to XenApp down the road, if we do see the need after the RDS pilot, is not rocket science.

Making a long story short, the environment is pure RDS to this day and has been working great down south. And getting ready to get it upgraded to 2012 RDS!

The lesson here is simple. A proper assessment of your SBC/VDI plan by someone that understands all this craziness is a must have and something that at the end of the day will cost you only a couple bucks, with the potential for saving several thousand dollars down the road.

For this reason we were the first ones to offer such packaged service, properly named TSEspresso (that you can read more about it here, http://www.wtslabs.com/tsespresso.html). To give you a fair and correct assessment of your SBC/VDI plans, with quick turnaround and on the cheap.

It is not Aspirin but guaranteed it will save you from headaches with your SBC/VDI plans.

Cheers.

CR

970 total views, 1 views today