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 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 live in the wild.

Ladies/Gentlemen,

Just a quick post to show you RDS-O-Matic in action. I just updated it to match what we have on the book I am releasing with Freek Berson. As I mentioned before this little handy Excel spreadsheet (Windows app will be ready and released at E2EVC Dublin) right now does the following:

  • Creates a complete RDS deployment with:
    • 2 RD Connection Brokers
    • 2 RD Web Access servers, load balanced using NLB on port TCP 443.
    • 2 RD Gateway servers, load balanced using NLB on ports TCP 443 and UDP 3391.
    • 2 RD Session Host servers with the desktop experience loaded.
    • 2 RD Licensing Servers, activated and set to whatever you choose (Per User/Device).
  • Creates a collection with the 2 RD Session Hosts on it.
  • Creates a test published application, WordPad, so you can try the environment immediately.
  • Enables HA on the RD Connection Brokers. For that you need to have SQL ready to go, the folder where the database will be stored has to be created and permissions for the RD Connection Brokers set. This is all detailed and explained in the book. We even cover the actual SQL install.
  • Retrieves the certificate (you enter the location for the certificate in PFX format) and deploys it to all roles. Here I assume you will be using a single, wildcard certificate for all the roles. If not the case, you can easily change the script you get to use different certificates.
  • Final warning, you must have the output folder created before you hit the ‘RDS-O-Matic’ button (i.e. C:\RDSOMatic).

Now that I am back at developing for iOS, is there any value in creating an iOS app that you can enter all this info and it will email you the script ready to be deployed?

Let me know.

Cheers.

CR

9,346 total views, 1 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

RDS 2012 and Remote Control. A workaround is found…

As most of you know, when Microsoft shipped its best RDS build to date (Windows Server 2012), it had to screw up something. It is part of the Microsoft culture. Deliver something good but make sure you screw up something.

I am ok with the screwing up part but I would appreciate people telling me in advance I am going to get screwed. The only person that can indeed screw me with such short notice is my wife AFAIK.

As Microsoft does not fall into the “My Wife” category they should have told us Remote Control was going to be removed from Windows Server 2012. Nope, they did not say anything even to us the RDS MVP Gods.

So after the Remote Control fiasco I decided to find a way to get at least some sort of half ass solution for the problem Microsoft created. Thanks to a great session this week here in Seattle during the MVP Summit delivered by the Multipoint Server guys, I realized we could come up with something.

Before you attempt to reproduce what was done I must clarify a couple things

  1. I am no licensing monkey. I know nothing about how the Ghetto Remote Control (as I coined it) is supposed to be licensed. Plus I do not care.
  2. It is almost certainly unsupported and if you call Microsoft and tell you are doing that, you are on your own. And you may get in shit as well. And if you tell Microsoft you read that here on my blog make sure you understand you are going to hell.
  3. I am not supposed to be liable for your actions. Again this is a Ghetto hack and unsupported. So you break it, you fix it. Not me.

All that said, here you have what I tested and know it works. Oh if you do not know what MultiPoint Server 2012 is, Google is your friend.

What do you need:
– MultiPoint Server 2012. If you have MSDN/TechNet, it is there. I have tested with Premium but given what we are doing I would assume the regular one will do the trick. I may be wrong though.

How to do it:

  1. Ideally create a VM somewhere and give it 2vCPUs and 2GB RAM (more if you want it to work faster). Boot off the ISO you downloaded from TechNet/MSDN. It is just a regular Windows install really.
  2. Once the server is installed, go to one of your RDS Session Hosts servers and browse to \\your_multipoint_server\C$. Under Program Files\Windows Multipoint Server you should see a Connector folder.
  3. Copy the whole folder locally anywhere on the RDS Session Host box and run the WmsConnector.exe file as Administrator.
  4. Just follow the wizard that is really simple. This will install the MultiPoint Server connector on your RDS Session Host. This has to be done to all RDS Servers.
  5. Once you are done, logon to your MultiPoint Server and launch the MultiPoint Manager. You should see the following screen.

    MultiPoint Manager 2012
    MultiPoint Manager 2012
  6. On the right hand side corner you have the “Add or remove personal computers”. Click on it. On the screen that will follow you should see all the RDS servers you have where the connector was installed. If you do not see them (i.e. they are on a different subnet) simply enter the IP address where it says “Personal Computer name”  and click “Manually Add”. If you see them, well then simply click on the “Add >>” button.

  7. Once you did that with all RDS servers they should show up under “Managed Personal Computers”  as you see on the screenshot above (where you can see my server named iqb-2k12rds and with a “Connected” status.

You are done.

So now, how do I remote control users? Still on the MultiPoint server, simply launch the MultiPoint Dashboard. The following window will show up:

MultiPoint Dashboard

Right there you will see all the users connected to the RDS Session Host servers (admins are NOT shown). Simply click one of the users (you can actually see a “live” view of the session) and on the ribbon at the top click on “Take Control” (or right-click the session and select it on the menu that will appear). The user should get a message and once he approves you are now remote controlling the user.

It is worth mentioning that a chat window shows up on the user session so you can actually chat with the user. Yes, a feature from the 90s delivered to you by Microsoft 22 years later. Well better late than never.

But wait, things get better. Another cool feature is delivered with the “Project” button. You can “send” your screen to all your users what is useful during April Fool’s day. Seriously this opens up a lot of cool scenarios, especially when thinking about training.

Want to limit which sites all your RDS users can visit with one click? Yep it is there under “Web Limiting” and its “Configure” button. Add the sites you want to allow/block and click “Start”. Simple eh?

Before you bitch about this solution, let me remind you again:

  1. I am fully aware this is half-ass and it may have a cost associated with it. As I said the licensing monkeys may know what that means in dollars. I do not.
  2. I told you before this is Ghetto style IT solution a.k.a. typical half-ass IT delivered stuff.
  3. And unsupported.

Even with all the above I still think GRC (Ghetto Remote Control) is better than what Microsoft delivered with Windows Server 2012 RDS.

I will keep digging deeper to see what else I find on MultiPoint Server that we can use to make RDS 2012 better… So back to work.

CR

 

40,478 total views, 2 views today

Windows Surface + XenApp + RDS 2012

As part of my work I try as much as I can to be up to date on all the platforms and its clients out there in the VDI/SBC space. That said of course I did have to get a Windows Surface tablet.

Before going a little deeper on my review, let’s clarify a couple things. This is the ARM based device (what means no Intel CPU inside) running Windows 8 RT. It is basically Windows 8 as you will find on any other device that runs it but as it is not Intel based, its software has to be compiled to run on the ARM CPU. Secondly, Microsoft has tighter controls on this device than on its sibling (the upcoming Windows Surface Pro) what means you can only install apps that come from the Windows Store.

Before you bitch about that, keep in mind the iOS devices are no different. You can only, officially, install apps from the Apple AppStore and whatever you can run on your Mac (running OSX 10.X) does NOT run on the iOS devices as they run on a different CPU than the Mac ones (Intel at this stage). So pretty much the exact same scenario with the Windows Surface device.

So what have I tested so far? A couple things:

– Offline usage (running locally installed apps, varying from games to real office ones)
– Accessing hosted apps on Windows Server 2012 (RDS Session Host with RemoteApps installed) over RDP8.
– Accessing hosted apps on Windows Server 2008 R2 with XenApp 6.5 installed over ICA/HDX.
– Accessing hosted desktops running Windows 7, hosted under ESX 5.0 with XenDesktop 5.6.

For the Citrix stuff you need the latest Citrix Receiver for Windows 8 RT. It is on the Windows Store. Contrary to previous versions this one does NOT require Citrix StoreFront and DOES work with Citrix Secure Gateway (what was a big surprise for me). Here you have a screenshot showing the Citrix Receiver on the Microsoft Surface:

Citrix Receiver for Windows 8 RT

Performance wise it did work perfectly when connecting to all the scenarios above. With RDP8 I was even able to run some WAN scenarios using my Apposite box (the mighty LinkTropy Mini) as you can see in the following video:

Windows Surface – RDP 8 WAN Testing

So what do I think about the Windows Surface as a potential iPad replacement? Well there are some MAJOR advantages on it and let me explain all I can see:

– Windows Surface runs Microsoft OS/Software. That means a couple things. First, RDP support is simply unmatched. That means whatever Windows Server 2012 with RDP8 delivers, it is there for you. The same can be said of ICA/HDX. This is for sure a big thing if you are after the best experience possible when accessing remote servers/desktops.
– Office. No half-ass support for Office docs here. The full blown Word/Excel/Powerpoint apps are here and I can tell you they do work PERFECTLY. Sure I had to get some updates for these but now everything seems stable and fast. And again, FULL COMPATIBILITY with Office docs. That alone is for me the biggest advantage going right now for the Windows Surface. I tried pretty much any piece of software known to man that is available for iOS devices to deal with Office docs. NONE were able to render all documents I had  100%. The Surface did it out-of-the-box.

The main issue right now for sure is the lack of apps if compared to the iOS ecosystem. If Microsoft can indeed convince developers to step up their game and start pumping out tons and tons of apps for the platform, I would definitely say the Surface can potentially rise as a very good competitor. And so far, as an RDP 8 endpoint, nothing can come close to it what is indeed impressive.

CR

1,353 total views, no views today

Intel buys Neocleus. More to come.

Two weeks ago I posted about my thoughts on the McAfee’s acquisition by Intel. If you did not have time to read it, it is here.

Resuming all that, the point made was that acquisition was probably just the tip of the iceberg and more than that, it would make a LOT of sense for Intel to acquire a company with a Type-1 Hypervisor offering like Virtual Computer, Neocleus, MokaFive and so on. I went one step further and explained on another post why Intel buying Citrix would make even more sense.

So as of this morning, you all heard the news. Intel is acquiring Neocleus, what gives them the Type-1 Hypervisor and the management solution in one package. With McAfee, that takes care of security.

The key question now is very simple. Is Intel doing all this (and again, there IS more to come, just stay tuned) to down the road build a platform that partners like VMWare, Citrix and Microsoft can continue building/improving/enhancing their virtualization solutions (mainly Hypervisors on this case) OR the plan is to come up with Intel’s own virtualization platform given the simple fact they already own the underlying HW one?

Again, the same way Microsoft leverages to its own advantage, the fact they OWN the underlying OS where everything runs, will Intel do the same from now on? A platform where everything runs faster and is way more tightly integrated to the HW layer below? Where all sorts of HW pass through work perfectly? Of course only when your virtualization platform is Intel’s. If you go down the ESXi, XenServer, Hyper-V, etc you lose all that. Is this coming?

To better understand what will happen in the next year, two little things the industry must keep an eye on: Intel buying NVidia (or a similar company in the graphics space) and how Microsoft and VMWare (Citrix as well but IMHO at a lesser extent) will react to all this in the short term (6-9 months). This will clearly give us the clues we need to determine if Intel is doing all this to make everyone else’s life easier OR to come up with its own, I-am-better-than-you-all platform that will keep everyone else out.

Time will tell if I am right or wrong.

So far I do not look that bad.

CR

1,107 total views, 1 views today

Intel buys McAfee. Does it matter?

As you all have heard today, Intel bought McAfee. And will pay well for it.
So the question now, at least from my end, is how this can possibly affect or help the virtualization market, the one you and I live and breath on.
First of all, we all know for VDI to take off as a mainstream solution (what it is NOT at this day in 2010 – and will not be for a long time) it must get cheaper. By cheaper it means being able to cram more instances per server. This can be achieved in several ways like using the latest, greatest and fastest CPUs you can get and with as many cores as possible (and of course using quick ass disk subsystems like Fusion-io, caching/dedup like iLIO, using tons of RAM, etc).

The point is on the CPU space there is nothing that prevents other vendors like AMD to get to the level Intel is at. In many cases in the past AMD actually delivered better silicon than Intel. So Intel needs a way to differentiate itself from their competitors. Bringing stuff like AV closer to the HW is one way of doing this. Good for Intel.

And of course getting this OUT of the VMs will for sure increase scalability. That was the reason why McAfee and others were coming up with appliances and lightweight agents (to run on the VM) to offload all that work outside the virtual environment.

The main question now is really how Intel will pull this off by not being a software company really. How they will get McAfee going. Of course I think it is just way too soon for any analyst to say anything about this. Historically Intel has not managed acquisitions like this well but they were never on such scale and with such reach like McAfee has (good or bad, they do have customers and a name in the industry, especially since that .DAT file fiasco that screwed up more computers in a day than any virus they were trying to protect).

In the near future I do not expect to see anything embedded at the HW level. This is for sure something that will come way down the road as you need to come up with something that can be leveraged to anything running on top of that HW. This means you either change the OS that will be running to benefit from these new HW extensions (like vendors did when using the virtualization components exposed by Intel and AMD on their CPUs) or, in this particular AV case, you get an agent running on the OS/VM. Not an easy task to do considering the amount of hypervisors and OSs now available.

That leads us to a very important thing. To minimize this and make things much easier, would not make sense for Intel to grab a Hypervisor vendor now? Given the three main players on this space now, VMWare, Microsoft and Citrix, I am sure the low hanging fruit here is Citrix and I even wrote about this ages ago on the post ‘Intel buys Citrix’

This would give Intel a huge advantage over any other company in the Virtual Wintel echosystem. Controlling the CPU, the Hypervisor that runs on it and extra features like AV, would give you the ultimate virtualization platform, where your solution runs better or has more features than anyone else. Example? All the HW fancy features are only exposed to your own hypervisor (like Microsoft is doing with RemoteFX, only available to Hyper-V hosts) and of course your hypervisor will scale much better than the competitors as you own and know it all about the underlying HW platform. Then the next logical step would be to acquire a graphics company like NVidia (as AMD owns ATI) and leverage all that into the platform, exposing it to the virtualization layer. Then, buy a good storage vendor and a management/layering one and they are all set.

Sure such scenario could potentially bring a lot of issues to Intel from a legal perspective, as it did to Microsoft when it became what it is today. But certainly it would simplify the virtualization market a lot (and yes, I know, locking everyone into their platform – what may not be a terrible thing as Apple has shown the sceptics with their iOS echosystem).

The bottom line is this acquisition for sure will help the virtualization space in the long run (do not expect mystical benefits happening overnight with this acquisition) but I see it as just the tip of the iceberg of what is potentially coming down the road from them.

Feds, you better keep an eye on Intel.

CR

935 total views, 2 views today

RemoteFX performance over lossy networks

Just a preliminary and quick video showing RemoteFX performance when loss is there, with and without our IPQ protection.

We have been testing it under several conditions, with different latencies and loss and will be publishing the full results soon. We also have some data on how much bandwidth RemoteFX uses. Just as a quick example, WMV-HD playback takes close to 30MBits; running an app like Google Earth around 9MBits.

My personal take after testing it, RemoteFX CAN be used over the WAN as long as you know exactly what your apps are (meaning, WMV-HD playback is probably a no go) AND also guaranteeing loss is minimized as it does suffer from it, making certain applications unusable.

CR

1,001 total views, no views today

RemoteFX First Impressions

As I did not have much time to test RemoteFX extensively, here are the first impressions of it and how we got it to work.

First of all, you MUST get a compatible video card. Not everything will work with Windows Server 2008 R2 with SP1 with Hyper-V, so you can get your Windows 7 VMs (with SP1 of course) working with RemoteFX.

I posted about it before. You can read the list of supported video cards here.

What did we get?

– HP desktop with a six-core AMD CPU and 8GB RAM.
– FirePro 5800 Video Card (also tried the unsupported Quadro FX 580 that by the way, does work too).

Initially I simply tested the Windows 7 VM connecting from the Hyper-V host itself but later got another Windows 7 SP1 box and used that one to connect to the VM.

Performance is decent I must say. I tried playing some Windows Media HD videos (make sure you disable multimedia redirection by using videoplaybackmode:i:0 in the .RDP file (save the RDP connection to the desktop and open it using Notepad). Also very important that you set the policy for RemoteFX (as I was not sure where to set it, I set it both on the client and on the VM itself). It is described here:

To set the experience index for connections using RemoteFX

  1. Log on to the client computer as a member of the local Administrators group.
    Click Start, and in the Search programs and files box, type gpedit.msc and then press ENTER.
  2. Navigate to Computer ConfigurationPoliciesAdministrative TemplatesWindows ComponentsRemote Desktop ServicesRemote Desktop Session HostRemote Session Environment.
  3. Double-click Set experience index for connections when using RemoteFX.
  4. Select the Enabled option.
  5. In the Screen capture rate (frames per second) box, click Highest (best quality), and then click OK.
  6. Restart the client computer.

The key thing to understand here is, why you may need RemoteFX. For example, during our tests, playing the WMV-HD tests, it used up to 30MBits so as you can see it is VERY bandwidth intensive. For comparison, running Google Earth in DirectX mode used around 9MBits. So basically the bandwidth will of course depend on the application being used. The same for how intensive CPU/GPU utilization will be.

I would expect applications like AutoCad to use way less bandwidth than something like WMV-HD and what we will be testing next is actually using RemoteFX over a typical home (cable/DSL) connection, simulated in our lab. By typical I mean a 10MBits down/1MBits up with 40-50ms latency and some packet loss probably in the 1% range (or a little more due to bursty loss). Given the first results we have seen, I am confident RemoteFX can indeed work over the WAN (at least bandwidth wise) depending on the applications.

Yes, before Brian Madden sends me a tweet or leave a comment here saying ‘MS says RemoteFX is LAN only’, I still want to make the point that IMHO, anything that is LAN only has its fate determined already. DOA. See my post about this here.

And still on the performance side, what we have seen in a nutshell is this: RemoteFX does work great BUT it is NOT the same as local. Simple things like Flip3D (using Windows key + Tab) are NOT as smooth as running them locally. Even Google Earth (that works just fine by the way) is NOT as smooth. But they both work and work fine, considering you are over RDP. For a BETA release we can expect it will be tweaked and improved even more before it hits the market.

As a sidenote, keep in mind there IS a bug on SP1 that throws a message on the RemoteFX event log about CPU encoding being used for ATI cards. It is a known issue and has been fixed apparently on later builds, what of course I have no access.  But for 1 VM testing like we did (I am after experience testing and not scalability – I will leave that to people with more time and resources on their hands like Ruben and Benny 🙂 ).

As soon as I have more results and some nice videos to show RemoteFX, I will post these here.

CR

4,472 total views, 1 views today