World of EUC

Your Teams is working Optimized in CVAD? Ours is still flaky as all get out. I was 2 for 3 today in successfully attending meetings. Even after TC reboot still not connected to Citrix HDX. What's the secret??????  :slightly_smiling_face:

Its a little flaky as well. The only difference it I dont have vGPU and no thin clients. If that is what you have still?

Why stop there? Run BlueJeans from within a Teams screen sharing session using Skype.

The craziest thing I’ve seen is the video ending up in a completely different citrix session to the rest of the meeting, so yeah, it’s still flaky

Haha <@ULC9J5XPB>  that happened to me today. Opened teams, opened zoom in teams, zoom opened another session and I was like wait where is things. To much man.

It might sound crazy launching Zoom from within Teams, but that’s the point. Teams is way more than just A/V, W/C and MS want it to be your single pane of glass for all collaboration, so if you want to use Zoom for W/C go ahead, but start it from Teams - make sense really. As to whether Teams or for that matter Zoom are ever going to be as good on CVAD compared to a Windows endpoint, imho the answer is no, even with a vGPU, simply cos it has many more moving parts, the features they support lag behind the Windows endpoints and Teams in particular can be a resource hogger so when they come up with these “Crazy” ideas you’ve always got to be thinking will this work optimised for VDI, if not what’s the user experience going to be like.

That is a good response Nathan. I agree with you

I will never expect videoconferencing software to be 'as good' in a virtual environment as opposed to physical, personally. Being able to provide a good experience overall is the goal here, especially with how many people are now working from home.

We might never expect it, but try telling that to the users, in their mind (probably rightly so) this virtual desktop should behave as a physical one. We know there is more to it than that, but it’s a hearts and minds game when using a virtual desktop solution. Published apps is a different story, but they too come with some user confusion. Oh well pays the bills as I always say.

I know over the last 6 years or so, alot has improved on this topic. I wish it was the same, There are to many variables that play to be honest. There has to be a better way to do it. Although I know there is alot of work that goes into it from a Developer standpoint as well.

I have that battle all the time, I'm afraid. I've even had to create _multiple_ PowerPoint presentations for leadership outlining VC in Citrix.

It's their own doing, most of the time. Find a solution and stick with it. Having to support Skype, Teams, Webex and BlueJeans is bad enough, especially when the latter two have no optimization for anything other than W10 endpoints.

I swear we only have BlueJeans because somebody in C level think's it's fancy and required.

The biggest issue I deal with is users using the user profile installs, Which we have to cleanup first before Zoom or Teams will work the way its designed. BlueJeans we are moving off of now. Which makes me happy.

Ray, are you using server 2016? or 2019? Do you optimize your image? Our Teams experience on 2016 is very poor. We get lots of complaints about it. Users constantly having to restart to "pick up" HDX optimization, which may or may not work.

We are using 2016, and I use Citrix optimzer. Lets do a Teams meeting togther so I can show you how it works on my side.

image.png

Then I delete these
				Active Setup looks for values called StubPath in the [GUID] subkeys under these two Registry keys –
				HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Active Setup\Installed Components
HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\Active Setup\Installed Components

I know you may have said this already. Are you using the Profile Container on Teams? If so do you have the exclusions in as well?

Also is the vGPU in place here from the XA servers?

I am using the verison as well :You have Microsoft Teams Version 1.3.00.28779 (64-bit). Citrix HDX Optimized".

Thanks Ray. We use the Citrix template as well, along with JG Spiers, and some additional settings that we have found. I'll send this over to my Citrix Engineer. We are on Teams version 1.3.00.34662. We do have Teams split out into ODFC from Profile container, and we are currently testing (i.e. my engineer and I are testing ourselves) putting Teams back into Profile container, and we do have the exclusions set. We do have vGPU, except we've turned it "off" via GPO (default hardware video adapter).

We have a case open with Citrix, and are working to recreate the problem, as we now have CDF traces running within the VDA and on one of the Microsoft clients.

I used to use Spiers script, But for us it was to aggressive. I cant remember the details. Most of it worked ok. But I had some new guys that couldnt understand, to check it before just running it.

I tend to stay behind a couple of version on teams. I need to look and read on what the new stuff enhances and fixes

FWIW, I had an issue in Teams 1.4.0.2879 where it would not allow me to set the default IM provider manually. I would check it in Teams, exit Teams, and the next time I opened Teams the option was unchecked again. Went back to 1.4.0.2781.

```&lt;?xml version="1.0" encoding="utf-8"?&gt;
&lt;LayoutModificationTemplate
    xmlns="<http://schemas.microsoft.com/Start/2014/LayoutModification>"
    xmlns:defaultlayout="<http://schemas.microsoft.com/Start/2014/FullDefaultLayout>"
    xmlns:start="<http://schemas.microsoft.com/Start/2014/StartLayout>"
    xmlns:taskbar="<http://schemas.microsoft.com/Start/2014/TaskbarLayout>"
    Version="1"&gt;

  &lt;LayoutOptions StartTileGroupCellWidth="6" /&gt;
  &lt;DefaultLayoutOverride LayoutCustomizationRestrictionType="OnlySpecifiedGroups"&gt;
    &lt;StartLayoutCollection&gt;
      &lt;defaultlayout:StartLayout GroupCellWidth="6" xmlns:defaultlayout="<http://schemas.microsoft.com/Start/2014/FullDefaultLayout>"&gt;
        &lt;start:Group Name="Customer" xmlns:start="<http://schemas.microsoft.com/Start/2014/StartLayout>"&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="0" Row="4" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Office 365\MS Outlook.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="2" Row="4" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Office 365\MS Word.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="4" Row="4" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Office 365\MS Excel.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="2" Row="0" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\ERP1.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="0" Row="0" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Zubehör\Explorer.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="4" Row="0" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\ERP2.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="0" Row="2" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Google Chrome.lnk" /&gt;
          &lt;start:DesktopApplicationTile Size="2x2" Column="2" Row="2" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\OfficeSuite.lnk" /&gt;
		  &lt;start:DesktopApplicationTile Size="2x2" Column="4" Row="2" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Office 365\MS Teams.lnk" /&gt;
        &lt;/start:Group&gt;
      &lt;/defaultlayout:StartLayout&gt;
    &lt;/StartLayoutCollection&gt;
  &lt;/DefaultLayoutOverride&gt;

  &lt;CustomTaskbarLayoutCollection PinListPlacement="Replace"&gt;
    &lt;defaultlayout:TaskbarLayout&gt;
      &lt;taskbar:TaskbarPinList&gt;
		&lt;taskbar:DesktopApp DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Zubehör\Explorer.lnk" /&gt;
		&lt;taskbar:DesktopApp DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Google Chrome.lnk" /&gt;
      &lt;/taskbar:TaskbarPinList&gt;
    &lt;/defaultlayout:TaskbarLayout&gt;
  &lt;/CustomTaskbarLayoutCollection&gt;

&lt;/LayoutModificationTemplate&gt;```


That shouldn't happen, I changed the path to an executable in our last monthly update and it works fine

What if I changed the path to the EXE, by using GPP shortcut replace?
Name and path stayed the same, just the path to the exe changed.

I had to take (2016) out of the path to one of my shortcut files, and it worked OK

I just restore my situtation, I destoryed it testing an send you a screenshot

Ok to cleary specify What I'm seeing, I just restored the FSL VHDX from yesterday.
I changed only to path to the EXE via GPP shortcut replace
I logon, startmenu looks perfectly fine. The update LNK files in the startmenu work perfectly fine.
BUT when I press one of the two affected tiles, nothing happens.

And this it what it looks like after the 2nd logon

Very odd. I'd create a whole new XML file and reapply it

Sorry for the question:
But what would be the difference, to create another XML with the same content?
Ps.: I apply it via GPO

Nothing, but it should work perfectly, so I can only assume something appears wrong with either the XML being applied or how it is being applied

I did as you asked, This leads to an interesting result. The missing buttons work again, but the Google Chrome vanishes.

Me too. This gives me headaches. This is the last tiny piece before this project goes into testing.

So I restored my FSL VHDX again from yesterday. The ERP icons now work. I don't know why, the Paths are literally the same.
I didn't change anything in regards to the Google Chrome LNK, the tile is still gone.

So ähm James, back to my original question. Do you happen to know how I can soft reset that, without deleting the VHDX?
Something I can hit with a logon script for example?

You tried deleting the CloudStore entries out of the Registry?

Is it this one?
```Remove-Item 'HKCU:\Software\Microsoft\Windows\CurrentVersion\CloudStore\Store\Cache\DefaultAccount\*$start.tilegrid$windows.data.curatedtilecollection.tilecollection' -Force -Recurse```

Yep, but everything from CloudStore downwards

Looks like you already have :slightly_smiling_face:

so ähm
`Remove-Item 'HKCU:\Software\Microsoft\Windows\CurrentVersion\CloudStore'  -Force -Recurse`
Like this?

Yep, I'd try that first. I'm sure there is a filesystem entry in LOCALAPPDATA\Microsoft\Windows as well but not sure it writes anything there

After the PS command, I'm back with working ERP tiles, but still a missing Chrome tile.

I renamed the LNK &amp; tile to just 'Chrome' but no difference

dumb question here, but why are all the shortcuts pointing at %APPDATA%, when I've used .lnk locations in my XML previously I've used %ALLUSERPROFILE% for most of them including Google Chrome, the only exception was for "File Explorer". Also MS don't recommend using an .lnk file, but rather the DesktopApplicationID, though I have seen that cause problems too.

No dump question!
Because I have this GPO active: `Remove common program groups from Start Menu Enabled` 
And create the startmenu via GPP shortcuts per user instead.

I’ve only seen this with Chrome, might be different versions in diff Master Image builds. Just drop Chrome for Edge ;-)

&gt; Also MS don't recommend using an .lnk file, but rather the DesktopApplicationID, though I have seen that cause problems too.
Well that's just the result of the Export-StartLayout command, I simple sticked to the Export

<@ULJ3D468M> Yes I've a hard feeling that this is because of Chrome. That's why I tried renaming it to something else, but that didn't help.

&gt; Just drop Chrome for Edge :wink:
Well I try it just for the lolz

Ps.: The Taskbar button, which has the exact same LNK path, works always!

In order to get Chrome to show a custom shortcut properly, we had to delete the chrome.VisualElementsManifest.xml file in the image

ohhhh so the problem is that the Chrome shortcut is not the original, but created by GPP shortcuts?

It might be the result, but it's not recommend cos file paths and shortcut names can change. Try changing the Chrome one to use a DesktopApplicationID="Chrome" see if it makes any difference. Or don't bother with any custom start layout, use the default and be done with it. MS have been known to change the goal posts regarding the startlayout, so I try and not touch it anymore

&gt; Try changing the Chrome one to use a DesktopApplicationID="Chrome"
With which command or tool, do i get the values for this parameter?

<@ULJ3D468M> <@UL5TCSW3D> Well lol, that worked fucking great on the first try.

Edit the XML and replace           `&lt;start:DesktopApplicationTile Size="2x2" Column="0" Row="2" DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Google Chrome.lnk" /&gt;` with `&lt;start:DesktopApplicationTile Size="2x2" Column="0" Row="2" DesktopApplicationID="Chrome" /&gt;`

I did understand you, But I would like to know where to get the ID if I want to link something else than Chrome, for example paint or mstsc

So it has to be a Chrome "issue", i play with the XML, google-search gives some interesting hints

<@UL5TCSW3D>
&gt; we had to delete the chrome.VisualElementsManifest.xml file in the image
This sound so promising, but it doesn't let the tile re-appear.
Google search also brings several hints regarding this file.

&gt; Just tested again, a FSL reset works..
Only for the first two logons. :angry:

I now decided to ditch the Chrome tile.
But now I'm back to the start, now the ERP tiles are invisible, when I logon with the profile from yesterday.

Hi <@ULHA2J7RT> I feel your pain with Win2019 Start menu issues:disappointed:. Can I ask, where is the xml stored? Local disk? Network share?  

Also I read an article online that best to apply the startmenu XML via a local GPO rather than domain GPO. 

Currently on a share, applied via domain gpo

<@ULHA2J7RT>: Create a GPP item to publish my script? I've been happily using it for exactly this purpose for almost 2 years:

<https://github.com/getvpro/Reset-Win10StartMenu>

Test it in your environment first, obvz

<@UMH935X7H> sound promising!
Tested on <tel:2019 180|2019 180>9 ?

I had originally written it for Win 10, I tested on Win 2019  1809 just now, and the TILE area reset as expected, but the pinned items in the taskbar area did not! I'll keep testing and upload a new version of the above script ifI can fix it

I already "fixed" several things for me personally. Testing it in the last few hours, it works very well so far!

I translated it, added a SIlentContinue to the Copy step and created an EXE with ps2exe
I will add the EXE to the start menu, so users can self service reset their start menu

Reset-StartMenu.ps1

<@UMH935X7H> Just to give back, I don't know if you are interested in my changes.

Thanks! I noted I didn't include if (test-path) statements on the original script, yeah :stuck_out_tongue:

Do'h you are right. That is much better than SilentlyContinue, I completely forget. But non-the-less, this was of great help!

Excellent! Glad to hear it I've uploaded a new version as well with the code stop prevent false errors: <https://github.com/getvpro/Reset-Win10StartMenu/blob/master/Reset-StartMenu.ps1>

<@UMH935X7H> <@ULHA2J7RT> Thanks for the thread and script. 
The script has resolved a really annoying issue with the start menu that I’ve had for a while. :+1::skin-tone-2:

Never ending story. Trying to rip out Google, but customer still wants it. Every time Google is updated in the Master Image (Evergreen) the icon breaks. Heck, even unpin repin does not work, icon okay on Start Menu, but not as pinned

nothing special needed, you don't have to clone and disjoin from the farm and reset the local host cache or anything.  Just a simple shutdown/clone, in an offline fashion, remove from domain, sysprep, rename, add network, join domain, and add to Citrix catalogues.  I'm old school and still use sysprep =)

just rejoin it to the domain with a new name

I haven’t used sysprep in forever, it’s really not needed anymore

If Mark Russinovich says we need it still, then we need it, haha.

the one where he says duplicate SIDs don’t matter and you only need it to depersonalize the image?

yeah, i was just making a joke though.   for VDAs sysprep probably isn't needed, but what about SQL severs or Domain controllers?

right...but from a lot of people just create a 'template'

for a blank SQL server it’s probably ok, I wouldn’t do it for a DC

yeah, just saying a lot of people have 'templates' that they create machines out of, that is just a VM that they do soem patching on, then convert it to template again.  Any VMs created from that template would have some SID type duplication.  So if you created a new domain controller from that...that wouldn't be good

I also sysprep. because of the machine ID. It will stay the same in my experience. Just had it happen.
Had issue with KMS. Some machine ID would act up. We used KMS on presistent machine. Not really smart I know.

Make sure u can get in with local admin. I creat a second on just in case
2. Clone
3. Leave it will with no IP
4. Log in
5. Sysprep
6. Setup.
But that was my old way I did it.

what is the issue with it being the same though?

Cloning in XA7_OS Tuning.docx

what do the nic settings help with? I have so little nic traffic on my vdas I never messed with it.

I think it was me tunning the nic back in the day. It's not needed any more. It was a old doc. Sorry I was trying to show my cloning steps I used to do.

Image from iOS.png

Image from iOS.jpg

Got some good info from support I will pass on. This option dynamically creates a local user on login. Works well, but doesn’t help me with my B2B users 

How do you provision the groups (user portal or request)?

they're also using App Layering, so I suggested to try updating VDA CU1 -&gt; CU2 to see if that would fix the issue.. if not then they could try to disable the driver signature check using bcedit (os layer) to verify it's about the signatures

FWIW, I'm on 1909 running 1912 CU2 with App Layering, and have those same updates applied, and don't see the issue. I went from base to CU2, though, so I can't speak to CU1.

Capture.PNG

I got this same issue the other day and it was a permissions issue on the executable.

thanks Brian.  Perms on the vuemappcmd.exe or the command line app reference?

Mine was a different app and it was inheriting permissions. I removed the inheritance and have parkinson to the users exclusively

VAD reboot schedules.....we have them all working as expected but seems to take longer to reboot sessions than the time frame we've set . possibly we're being too aggressive for our time window but sessions seem to go down by only about 3 or 4 every 10 min which seems a little conservative.  wondering if we have any more control over it (we're Nutanix for our hypervisor for reference and set the max concurrent actions to 100)

The old "blacklist versus whitelist" argument.......

because BCR doesn’t work for every website, tbh there are quite a lot of sites where it simply doesn’t work.

If you quit and restart Teams, does it enable? We're having similar problems, along with Teams losing HDX / connection during (or between) meetings, and users have to quit / restart. Sometimes that works, other times user have to log off and get on a different server for HDX to reconnect. We've had a case open with Citrix for weeks with no real movement. We are running 1.3.00.34662 on W2016 (1607) and VDA 1912 CU2.

Are you using any optimization for your image build?

Thanks Tim. Yes we do have some optimisations, mainly based on George Spiers and other Citrix best practices. We also use 1607. I'll have to do more tests on quit/start Teams, it is new for me, I didn't even face the issue yet. Are you using FsLogix? We don't, we use UPM and it works great for the Teams loading part. We are a large enterprise and we have MS overseeing the project so I'll make sure to raise those issue(s), it should get the proper attention.

Interesting - we use the George Spiers optimizations as well, and I'm wondering if that is part of the problem...Yes, we are an FSLogix shop, splitting our profiles and Outlook data. Teams is currently in the ODFC, although we are testing it in profile container with no real conclusion yet.

Interesting indeed! It is a script we apply at boot time, I could exclude it quite easily. But we have also other optimization settings applied by GPOs, I'll review them but nothing too aggressive. 

For now I need a bit more data to see how often this does happen. Is there a way to find out by PoSh if Teams is optimized? it would help quantify the issue. 

I'll keep you posted if I find something significant.

I am currently in the process of creating a script to find out if Teams is optimized. It’s not easy though…

I found out there is a new Teams version available, 1.4.0.4167, I'll give it a try.

<@ULHA2PS2H> we’ve got exactly the same issue as you, same versions as well, no fslogix

Tom - are you using the optimization script from George Spiers as well?

Ok, we're using those as well, and put some of the optimizations from George Spiers into the Citrix Optimization template we're using. Which one are you using?

Have to check that tomorrow ...  you suspect any link with it ? (Btw, you also have incoming video issue is one joins a meeting if teams was not running, see my other post)

It seems to be the common factor....Other people with similar setups are working OK. Ours has been crap from the beginning. We've ruled out clients, as it occurs across Mac, Windows, and linux, across all major versions of Workspace App. I suppose it could be some Citrix policy somewhere that we all have configured...although that seems unlikely. Most likely to me is some setting we've all changed when optimizing our images. Currently, this is just a hunch, though.

Interesting indeed, Especially when you say other people with similar setups are working fine ... that gives some hope .... btw, I typically think quickly to antivirus stuff (trend in our case), reviewing exclusions here 

We do not have any AV in our image, and I can definitively say it's not at fault in our case.

No Citrix optimization for us, only George Spiers script with a few updates + few other settings by GPO. We use Symantec for AV.

Citrix optimismer 2.6.0.118, optimization xml “Citrix_Windows_Server_2016_1607” version 1.2 from 20/11/2019

Did you find a fix guys? Did you raise a Citrix/Microsoft case? I had it only one time so far, log off/log on fixed the HDX offloading.

No fix for us yet. Logging off / back on *sometimes* resolves. Although that's not a solution I can send out to my employees. It needs to work as well within Citrix as it does outside, and we're just not seeing that.

yes absolutely it's not acceptable, we already have to do enough compromises (crappy install/user cache, more resources needed vs S4B, etc.) to get Teams working, it's enough.

I was trying to see if there is any case opened by anyone facing this issue so I can link my case to it when I open one. Let's put this way, MS wants us to be successful with Teams rollout but we are a Citrix house unfortunately for them :grin: 

Also don't quote me on this but we should have a better install/user cache in 18 months 
... at least there is a plan to address this! and so far in our pilot, Teams with UPM does a pretty good job, Teams is a bit slower to start  but we start it automatically in systray at logon so user xp is not bad.

We do have a Citrix case opened, and I think we have an MS case as well. Citrix case is: 80289484.

what we've noticed : Appcrash of WebsocketAgent.exe (event 1000), crashdump provided to citrix ... Anyone seeing same crash ?

Thanks Tim. I'll keep an eye on this Tom.

Tom - looking back through my logs, and yes I see that as well. I have ~20 instances over the last week.

The week prior to that we had ~35. Not sure if it's trending yet...

what version of DotNet are you? We use 4.7.1.

W16. VDA 1912 CU2. Symantec Endpoint AV. UPM for profile. FSLogix only for AppMasking in use but profile component is installed. XenServer 8.2 with XenTools 9.0.42.

W16, VDA 1912 CU2. No AV, no UPM. .Net version 4.8. XenServer 7.1 LTSR, with XenTools 7.1.1305.

No, FSLogix for profiles and ODFC, separated.

No Ivanti and WEM installed but not use anymore

FYI : About websocketagent crash: we have a private running in test for some time now without any issue, I understand it will be included in 1912 CU3 ...

Thanks Tom! We've pinged our escalation engineer and they are investigating this private fix for our environment. I hope to hear more next week.

Yep I get that. Only solution is to kill Skype

i always disable that service.... think high cpu load cause through no internet access?

amd epyc is very nice, but core licensing is not that good

but yes, we invest much time to choose the right cpu between cores - ghz and applications

Have not seen that in my lab. Upgraded that one also from CU1 to CU2.
The web.conf still contains the detection rules, can you check what's in them now?

Have not seen that, Have server ENV where I switched from CU1 to Cu2 without problems.

I am pretty certain you can't do anything like that via GPO. The language setting is stored in a JSON file.

Thank, Kasper. I managed to  get the windows regional and language settings working as desired (as per my other thread) - I reckon I'll get the IT team do some comms for how to update Chrome/Teams to suit.. just a smallish gig so it should'nt be an issue, cheers

Add the setting into the default JSON file as laid out in my latest Teams article. It might well work, but I haven't tested

Nice one, James - I'll have a look... have you ever had to add/push Chrome languages via policy/reg... the interwebs is lacking here too :slightly_smiling_face:

<@ULL6F7DRA> is probably more your man with Chrome and Edge Chromium, I haven't tried

I think the browser language follows the OS language, at least that's what I see with Edge (Chromium)

You can set the language based on groups, I added myself to a nethlands group in AD and my language in Teams and Sharefile changed

special folder redireciton is already prohibited

Are the affected users using EDT? Has a netscaler firmware changed recently? We hit a bug with a netscaler firmware upgrade and the symptoms were similar.

<@UMB4XQE79> you are sure there are no updates on client side?

And the message ,,invalid parameter" is on client side or from server side?

On the client side, inside the grey screen session

In my experience, every time I get a grey screen while brokering a ICA session. It 90% related to CTX hooks related to a application. Electron apps are a killer in my experience.

The problem is gone as soon as it came without changing anything :face_with_rolling_eyes:... all this stuff is getting me down. 

I’m getting grey screens today and it’s driving me insane

I even tried with the citrix quick launch tool, it sits on a grey screen for 25 seconds before the password prompt comes up

hi, i had the same issue yesterday. The customer added some system path variables and that breaks the ica initiating... after sorting that paths (testing removed) all works fine - maybe like that?

Launch regedit and attach the NTUSER.DAT from the default profile to see if those devices are listed in the printer connections key.

I'll take a 2nd look, I already did that, but you may never know.

Currently redeploying the image, because I can.

Yeah, worth searching for one of the printer names to see if it's in there somewhere.

Don't forget to unload the hive when you're finished!

Solved it partly.
One of the MDT User accounts was in a OU with GPP Printer preferences.
Moved the Service Account to a clean OU and rebuilt the image.

What's left is the following, and that's like impossible:
Fresh Testuser Session. No Network Printers. I map one Printer. Logoff. *Delete my VHDX*.
Logon -&gt; *Printer still visible*, but can't view it's properties, set it as default or delete it.
There is even an entry in the registry!

There is no Folder Redirection for AppData I can spot.

have a look at printer-enum registry keys: <https://social.technet.microsoft.com/Forums/en-US/6e264af4-192c-4b0f-9fff-f7cd0bb5655f/windows-2012-rds-printer-duplicated-and-ghost-printers?forum=winserverTS|https://social.technet.microsoft.com/Forums/en-US/6e264af4-192c-4b0f-9fff-f7cd0bb5655f/wind[…]rinter-duplicated-and-ghost-printers?forum=winserverTS>

So this seems to be my problem in one or another way.
Except that I can rebuilt the machine at any given time + a MCS reboot resets the VM, so all the workaround from this (and similar) articles wouldn't apply IMHO because why should I need to clean up registry keys, when the machine is still fresh?

Guys, not sure if anyone can answer here - can't find a solution or document anywhere online.

A new client wants to start using hello 4 business on corporate laptops, the user gets a certificate generated by internal AD, all works fine.  Now they want to integrate SSO to citrix... As far as I know, there are only two ways to do this

1. Use nfactor to look for the cert and logon if there
2. integrate citrix FAS to Azure AD connect and then saml logon using azure AD generating a second smart certificate linked to shadow accounts in the local AD domain
The client wants to use the local cert that they generate for h4b to do the SSO via fas, but as far as can see there's no way to do this?

Has anyone managed to get SSO working with H4B using FAS or not? As far as I am war, even Citrix Workspace uses FAS and azure AD connect to achieve this...

Thanks in advance..

It doesn’t work in Workspace (coming), but works for web (AAD) 

File explorer looks incorrect. This is what I use:
```&lt;taskbar:DesktopApp DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Programs\System Tools\File Explorer.lnk"/&gt;```

Am I not allowed to use any LNK I want?
I use the following GPO and built my shortcuts with GPP instead:
*`Remove common program groups from Start Menu`*

I want to say that I've tried using custom links to the same EXE in the past on my home machine, and it always opened in a new icon in the toolbar.

I could be wrong now, but it's worth a look.

So it worked for Google Chrome, but not for File Explorer. But that's because I deleted `%APPDATA%\Microsoft\Windows\Start Menu\Programs\System Tools` via GPP, to keep Start Menu clean.
I temporary disabled that GPP and then it started working for File Explorer. XML looks now like this:
```  &lt;CustomTaskbarLayoutCollection PinListPlacement="Replace"&gt;
    &lt;defaultlayout:TaskbarLayout&gt;
      &lt;taskbar:TaskbarPinList&gt;
		&lt;taskbar:DesktopApp DesktopApplicationLinkPath="%APPDATA%\Microsoft\Windows\Start Menu\Programs\System Tools\File Explorer.lnk"/&gt;
		&lt;taskbar:DesktopApp DesktopApplicationLinkPath="%PROGRAMDATA%\Microsoft\Windows\Start Menu\Programs\Google Chrome.lnk" /&gt;
      &lt;/taskbar:TaskbarPinList&gt;
    &lt;/defaultlayout:TaskbarLayout&gt;
  &lt;/CustomTaskbarLayoutCollection&gt;```

If you're set on hiding a lot of those items, you can always create App Masking rules to hide the ones you don't want.

File Explorer acts very weird when pinning it to the  taskbar via LayoutModifications, but also with WEM.

I usually leave the "default" pinned file explorer.

After the tipp from <@ULC9J5XPB> it seems to work well. (But this was only a very short test)
We'll see what the next days bring.

IE is the worst, IMO. If you delete the IE stubpath entry, the shortcut is never even created in the first place.

Fslogix for start menu management is so much nicer - deployments have been stable as ever since i started using it. Common start menu is now my best friend in deployments

We still like to set a default start layout via XML, and lock those items, but users can pin whatever they want. No issues yet, though I agree that coupling it with App Masking has made it a better experience. No more scrolling for days just to get to stuff at the bottom of the list.

<@UKJTJD4KA> So instead of hiding common programs, you tidy it up and use app masking to only present the LNK each user needs to see? Interesting, but doesn't help with my StartMenuLayout and TaskBarLayout Problems. Only makes my GPP approach for a users start menu nicer?!

I try and avoid user GPP as much as possible because of the effect it has on logons, personally. We do use WEM to assign certain apps to different groups, though I may look at moving that function to App Masking, too.

I have no WEM available, only at 1 customer out of 20. I :heart: WEM, and it works flawless since about 3 years at that one customer, but I had to find a way to provide custom start menus without it. So I was recommended hiding common programs + user shortcuts GPP

And 99% of it works. I just the missing tiles sometimes, and the taskbar layout, which I was able to solve yesterday thanks to you <@ULC9J5XPB>

Yup! <https://jkindon.com/2019/10/09/appmasking-the-windows-start-menu/|https://jkindon.com/2019/10/09/appmasking-the-windows-start-menu/> combined with custom start (including taskbar) - I just make sure everything comes from common start locations and bobs your uncle - except edge, which is a prick and dual icons no matter what you do 

I thought I read that old Edge would eventually be removed via an update? I hope that's true, because nobody uses it.

For Windows 10, yes, but old Edge isn't on Windows Server, isn't it?

<@UKJTJD4KA> I'll read it, but I'm not sure if I want to throw away what I achieved until now^^

Old (legacy) Edge is not available in Server OS

I’m not talking legacy - new awesome edge ain’t awesome with taskbar layouts 

<@ULHA2J7RT> come towards the light :grimacing:

I usually handle the pinning of "stuff" with WEM

Yah I’ve given up on edge in taskbar to be honest - stick it as a tile and exit stage left 

Yeah, either that or just leave it to the users, most are able to pin apps

Wait, Edge Chromium has issues pinning to the taskbar?

Agreed 100% - even self pinning of edge does dumb shit in my WVD deployments 

Hmm, I'll keep an eye on that when we finally make the switch. If I can make it work, I'll post what I did.

I know there are differences with pinned Edge in Windows 10 vs. Windows Server