World of EUC

Agreed, we have opened a Microsoft case for this. This is reported as known bug. For workaround currently you can disable safe links only 

<@ULLRRJRHC> Thanks - this issue was driving me mad. Any ETA on a resolution?

Currently not, heard this from a customer. If any update of this I will post it here 

Thanks <@ULNFS3KF0> for sharing the issue here

Do you mean a locked but active session, or a disconnected session?  Does it fail or establish a new, separate session?

I think locked and active, but I’ll have to check.

when the thin client locks, it sets the session to disconnected. This seems to be a citrix cloud/wyse thing because on prem it stays active

I think that's a Wyse setting, I vaguely remember coming across it at a customer with Wyse clients. We had an issue with Citrix sessions 'crashing' the thin clients, turned out sessions were getting disconnected and the Wyse device profile apparently has a setting to control what happens when a session ends and they were set to power off.

We have a few RDS servers that occasionally fail to shutdown and reboot during their schedule cycle. It appears it may be due to the infamous ghost session. Anyone using some sort of shutdown scheduled task script or automation to help clear these out before a studio reboot?

Were these persistent machines or non persistent?

Ah that’s weird - that should not be happening and you shouldn’t need any form of policy around account password resets like pvs 

No. Only crucial when you install teams for the first time. You should install the vda first and then Teams. When upgrading, a new installation is not necessary 

That's really on a per-app basis.  I would think the app needs to support it.

Is there an article covering step by step? If our app supports it, we'll implement it.

i would start with the app vendor.  "Does your app support SSO?  If so, how?  AD passthrough, SAML, etc etc"

fair enough.. Thats an internal application. I ll check with our developers then.

Yes, the app developers would have to query the OS for the creds or a token that validates it against there own credential store.  Nothing to do with Citrix or Windows

In Citrix we just have to create a traffic policy enable sso in it and apply it to vip. Is it the process to enable SSO for SSO compatible apps?

requiring telemetry on your usage, rather than 'optionally collecting'

I was just thinking what about lic server that have no oubtbound connection. Then I seen this. "This release enables automatic license telemetry reporting to Citrix. For Citrix License Servers that are unable to connect to the internet, we also offer the option of manually reporting licensing telemetry data to Citrix. To enable manual reporting, please follow this process: <https://support.citrix.com/article/CTX472950/how-to-manually-upload-telemetry-data-to-cis>"

No telemetry...no support
Enabling license telemetry is a CSS program term requirement for receiving product support

Frankly, I'm surprised that this hadn't happened sooner.  It's been so easy to use a Citrix license file from anywhere and just name the license server the same.  It looks like they are finally going to try and enforce licensing entitlements.

I get a 404 when clicking on the link in that article to the latest version of the license server. Is it just me?
Attempting to download it from the VAD 2212 page doesn't work either. Great.

Also find it interesting that the latest version is 111720-42000.  It was probably released in the time between when the article was initially written and eventually approved for release...

<https://www.citrix.com/downloads/licensing/license-server/license-server-version-111720-42000-for-windows.html>

From previous discussions with Citrix staff, they are aware of a significant amount of non-compliance with licensing and have been looking at technical license enforcement for some time - looks like they've decided to do something about it

I have a ton of defence customers with isolated networks who are going to love this :face_with_rolling_eyes: At least it's only every 90 days and not monthly
"Customers reporting manually need to upload the upoload_compliance*.zip files at a minimum of once every 90 days"

I'm surprised it took this long for this to happen as well. Will probably have to have some exceptions for certain black site customers that don't have any allowed egress

nah shouldn’t be an issue - I ran it for months without any issues ;)

I even did it with SF. I know it is not supported but it did work. Not to give bad advice

i don't want to do it but the client is leaving me no choice

Nevermind
<https://www.citrix.com/blogs/2022/06/22/planning-for-citrix-xenapp-and-xendesktop-7-15-end-of-life/#:~:text=Per%20our%20Product%20Lifecycle%20Matrix,Life%20on%20August%2015%2C%202022.|Planning for Citrix XenApp and XenDesktop 7.15 End of Life | Citrix Blogs>

Remember our friends in Adelaide who like wine? That was vsan, my biggest learnings from that one with persistent mcs provisioning was make sure you full clone so you can alter stuff when they double what they told you they needed on developer vms :flushed: and that they were focused on storage without thinking about CPU and Mem, and ran out of that quickly :joy: 

Ah yes so it was! For this one persistent desktops will be deployed via their existing provisioning process, but good to know :slightly_smiling_face:

We run vSAN at one of our shared-storageless remote offices.

Our vSAN guy says be careful of dedup and compression; only use it if you test thoroughly first.  It can destroy i/o.

Use the sizing guide, and try to go for for more than one disk group per host.

I have a feeling VMW will kill off linked clones just to push everyone to Horizon with instant clones.

We are an MSP and use VSAN everywhere, although we are rather small. I love it, everything runs very well, they are our best environments.

<https://support.citrix.com/article/CTX477614/citrix-license-telemetry-faq|https://support.citrix.com/article/CTX477614/citrix-license-telemetry-faq>

First thought on the top of my head is that it might be creating the users profile.  BUT, is this a multi-user OS and it only happens for the first user's first logon, or all users' first logon?  When you say fresh VM, is this a brand new persistant VM or a rebooted non-persistent VM (where the C:\users folder is blank on reboot)?  Roaming, local Windows Profiles?  Using UPM or other profile management?

Did the vast majority of my environment. I use the VDA Cleanup tool and it's basically a clean install after that for the VDAs. The core services went well except for not knowing that Configuration Logging needs to be disabled during the upgrade on the DDCs.
<https://support.citrix.com/article/CTX279999/upgrade-from-xendesktop-715-to-cvad-1912-ltsr-fails-with-exception-databaseconnectiondetailsalreadyconfigured|https://support.citrix.com/article/CTX279999/upgrade-from-xendesktop-715-to-cvad-1912-[…]s-with-exception-databaseconnectiondetailsalreadyconfigured>

The way I see it, we have two options:  1st, allow multiple connections to the same profile container.  I don't know if this is still/ever was the recommended config, open to input.  2nd, publish seamless apps and published desktops from separate delivery groups, each pointing to a different profile disk.  While this would work, I can see potential for confusion from the user (file exists in downloads/desktop folder on published desktop profile, but isn't there in the seamless app profile.

From the user perspective, I think scenarios like launching a published desktop from your workstation first thing in the morning to work out of most your day, but then using a seamless app from your tablet during a meeting, is a valid use case.

I use a dedicated DG for published apps.  Concurrent access to FSLogix containers was a nightmare for us.

I would take a look, at this article by :goat: Kindon:
<https://jkindon.com/citrix-upm-and-fslogix-containers/>

<@ULHA2NPPX> are you using folder redirection or anything like that to, for example, make the "desktop" folder consistent between both DG?

you can do that if you so chose.  Depending on the needs of your specific published apps

Published Apps are typically (not always, but a lot of the time) a prime candidate for UPM file based profiles - lean mean and quick. Combine with folder redirection for appropriate data (maybe documents, maybe desktop etc) and you are home free. Typically I see published apps on dedicated delivery groups - not always of course and it depends on the apps and use case

We do redirect my docs to home dir, and have always instructed users to keep any important files there vs the "desktop" folder, so I suppose that can be our "out" if someone complains about desktop folder discrepancies.  Good to know about UPM - I think in this case, if I can get away with it, continuing to use FSLogix for those and even avoiding folder redirection will be the simplest from a management perspective.  I'd say we're &gt; 90% published desktop users/sessions.  Good article, thanks for chiming in!

Only when you add new Machines to the mcs catalog

what about if you are changing the NIC...my colleague told me that you have to create a new catalog

Just create a new machine catalog, new non-persistent and join them to the delivery group. You're good to go then

anything you edit at the VMWare object level meaning CPU/Ram/Disk size after the machine catalog was deployed need to be edited on each machine. The only thing that will get update is things in the OS level. But as <@USCA04SN9> said if you redo or build a new catalog then you the machine catalog provisioning scheme will allow this.

Here is a example were I had to do something simliar
```##Records the Powershell transcript
##Import the VMware module
##Connects to vCenter, you can let it prompt you for username password or you can define it. Example(Connect-VIServer vcenter.lab.local -user Domain\username -password "password") 
Start-Transcript -Path "C:\Temp\transcript0.txt" -NoClobber
Import-Module VMware.VimAutomation.Core
Connect-VIServer vcenter.lab.local -user VSPHERE.LOCAL\Administrator -password "password-example"

##Finds the VDI with a name like LABDaaS and exports it to a CSV
##Imports the CSV file
##Does a Loop and shuts down each VM and enables hotadd for memory and CPU. There is a 60 second sleep between each VM. You can change this
##Starts each
$VMsToUpgrade = get-vm |?{$_.Name -like "*LABDaaS*"} | select name | export-csv -Path C:\Temp\VMsToUpgrade.csv
$vmName = (Import-Csv -Path C:\Temp\VMsToUpgrade.csv).Name
ForEach ($vm in $vm_name){
Shutdown-VMGuest -VM $vm_name -Confirm:$False
Sleep 60
Get-VM -Name $vm_name |Set-VM -MemoryGB 12 –NumCpu 2 –Confirm:$False
$vm = Get-VM -Name $vmName
$spec = New-Object VMware.Vim.VirtualMachineConfigSpec
$spec.CpuHotAddEnabled = $true
$spec.MemoryHotAddEnabled = $true
$vm.ExtensionData.ReconfigVM($spec)
Start-VM $vm_name
Stop-Transcript
}```


image.png

im using mine right now. Works great with my most recent laptop and IPAD switching back and forth. They will have to pry my X1 mouse out of my dead cold hands...

I won my in 2015 from a CUGC event, Still in the package. Have no idea if it works. But maybe one day it will be worth some $$$

haha. yea I have a second one somewhere I think.  I wonder what might happen if I called into support and tried to log a ticket for the X1 mouse..... :joy:

I remember there was a clear one. Instead of black.

hahah just caught the support question LOL.

yeah, i think the clear one was the prototype

now, if it only had a roller ball instead of a laser...

Still fun to search Ebay for 'Citrix'  <https://www.ebay.com/itm/175597134490?hash=item28e267a29a:g:KvEAAOSw7RFj2v1O&amp;amdata=enc%3AAQAHAAAAkMiUoCJEFY8t8LPZXEKwAMdklIAmCz0vLHBQWXIdS%2FzfQTgFr%2BxxqMvEhecFwCQ%2BB4LB7XYPWojLBWIhKiet0A3t5OpsfXtxOVi1Wu2CqVmdGExW0U3Gye0%2FQPz6Ijw3PWSModL%2BcTqCuLe%2Felyfi%2BX4kaB9J1FRIw8blQ73kLYvVZEv30md5pFWqkhnHvnZUA%3D%3D%7Ctkp%3ABk9SR6jAvKfGYQ|https://www.ebay.com/itm/175597134490?hash=item28e267a29a:g:KvEAAOSw7RFj2v1O&amp;amdata[…]w8blQ73kLYvVZEv30md5pFWqkhnHvnZUA%3D%3D%7Ctkp%3ABk9SR6jAvKfGYQ>

<@ULLH5JGFR> This one might make you feel better,  <https://www.ebay.com/itm/224415664684?hash=item3440374a2c:g:zlIAAOSwgQBgbf3J&amp;amdata=enc%3AAQAHAAAA4PSyu%2BlxJDkAM%2BqbgXw8B2%2B84P17S4IFWaTkpKJVycbpp%2BVwV3%2FdONxAY4PIjZVdAGANGxgKffxnRbJBoXQYSm7ndeM2iNbTHovw7Twj%2BQYCO%2FNMwAidFZoTpktjeAmyYvYV%2FnR9Ofo1SKUJyRk%2BYXRjZ2L%2BlBkNUZVYfa0phhenoaKg6b%2F9IIA1i7PfmuFN%2F4wlhljlpH8mHVNduiW3nQBTkTcI%2Bpjyu46LqPDPropuzumk%2Fe6d4T7HlkC4l8awrkuJDmh0228ppmp7Md%2Fdu4u0RBtqSNMJZUTf3P33EpkE%7Ctkp%3ABFBMqsC8p8Zh|https://www.ebay.com/itm/224415664684?hash=item3440374a2c:g:zlIAAOSwgQBgbf3J&amp;amdata[…]mh0228ppmp7Md%2Fdu4u0RBtqSNMJZUTf3P33EpkE%7Ctkp%3ABFBMqsC8p8Zh>

It doesn't surprise me one bit that <@ULN862X8E> has that.  :smile:

IMG_6533.jpg

I’m using the prototype every day, good stuff but it still is just a mouse, don’t get it why it should be special :rolling_on_the_floor_laughing: 

If you use it it helps with ICA latency hahah

That is a Joke just in case someone reads this and thinks I am serious

It’s using the TCP metaframe XP features which makes it a great device on any connection

It's the MF20.DSN file that's the secret sauce

Haha 10 bucks on the first one. It has scratches, but the second one is to much for a mouse with a sticker lol.

I had a clear one somewhere, but there’s no point now that ipad supports mouse input natively

figured this out...have to select Disabled instead of selecting "Use default value: Disabled"

I have a guide on Zoom in CVAD if you need it.

Hey, AVD is a platform service in Azure with its own mgmt tools in the Azure Portal / via PoSH... you can use something like Nerdio to close a lot of the gaps when comparing to Citrix (Studio/Director etc)... that said, you can use licensing to run Windows 10 multisession in Azure with Citrix / Citrix Cloud... but this is not 'AVD'... James K has a nice article here: <https://jkindon.com/citrix-cloud-and-avd-fluff/>

Server 2022 has slightly worse performance than 2019 - check out this article (it's from september 2021 so things may have changed since then)
<https://www.go-euc.com/windows-server-2022-performance-testing/>

Same goes for 2019 compared to 2016. And the same goes for 2016 compared to 2012 :slightly_smiling_face:

That said, 2019 looks easier on the CPU compared with 2016 in the above article... the main driver here is OneDrive for FileSync and Storage Sense so I'm leaning towards 2019 as that will do the job without so much tax on the resource...

EOL for 2019 is jan 2024. So less than a year. We run 2019 - no issues. But already planning to move to 2022. Keep good notes on issues/fixes as I will sure ask for them 8-)
And If you going to do any performance testing for user density - launch Teams for those test users.
Our current 2019 issues:
Searhindexer sometimes prevents users sessions from logout. (Doesn’t happen often - we keep checking sessions running over session limit and killing searchndexer there.)
RAM cache filling (also not something which is happening often, not often enough to find the root cause)
Couple of Windows Update related issues - these were cough during UAT and weren’t released to Prod - (start menu not launching, one drive not signing in)
VDA 2203 CU2 - significant increase in ICA Latency + Teams sometimes not turning on optimisation- we are in a proceed to rolling back to 1912 CU6

Nice insights, Vadim... extended support runs until 2029 for reliability/sec patches which is key... I'll need to weigh up the performance hit/cost of infra..   . Azure is on the cards for for the client so I see tis being required for 18 months to 2 years...

for 2019 the extended is 2029.  Kind of like Server 2012R2 is for Oct 2023

Yes, we see that occasionally, although it's not as bad with some updates. What OS and what version of VDA?

We saw that on VDA 2203 CU1 as well.  After reviewing a few different discussion threads about this, we ended up adding these two registry keys to fix it.

HKEY_LOCAL_MACHINE\SOFTWARE\Citrix\Ica\GroupPolicy "EnforceUserPolicyEvaluationSuccess"=dword:00000000
HKEY_LOCAL_MACHINE\SOFTWARE\Citrix\Reconnect "FastReconnect"=dword:00000000

Not only were we seeing the issue with it creating a new session instead of reconnecting to the existing session, when the original session would reach the Idle -&gt; Disconnect timeout and the session would disconnect, all the existing sessions on the server would lock up and new sessions would get stuck trying to login.   We could not restart the server gracefully, we'd have to force it to reboot.

We’re piloting CU2 now and we left the reg fixes in place.  I don’t believe this is fixed in CU2 but I don’t know for sure.

Oh boy. You’d think if this was a widespread issue Citrix would’ve at least included those reg keys with CU2. We are kicking off a large scale upgrade from 1912 to 2203cu2. I hope we don’t get bit :grimacing: 

<@U02TDPX0M5G> we are on server 2022 and VDA 2206. So I guess I will have to give the reg settings a shot. Strange if this is a bug since 2203 and not fixed in 2206. Maybe not even fixed 2209 och 2212 then...

Fixed in 2203 CU2 :i_love_you_hand_sign: If you're on current release I would go for 2212 as there are several bugs fixed. We had a huuugee discussion about disconnected sessions here.

<@U023HEDJVH8> 2203 CU2 this issue is fixed

<@ULCBDDJCR> but we have 2206, other issue or maybe not fixed after all?

Any idea why the registry entries were not written during the vda 2112 installation? I used the VDA removal tool, then installed VDA 2112. I'd rather not hack the registry on every new server if I can find the real solution.

My process after the install to seal it (they are PVS) is I run optimizer, autoruns, then the PVSSeal script.

Interestingly enough, after removing WEM 2302 and going back to 2212, everything works great. So, must be a bug with WEM2302, or at least some change that affects my environment.

Have you met my friend <@ULNE983N3>? Loves this stuff 

Review your Azure AD sign-in logs to see what’s going on and whether the sign-in is passing CA policies

Dave we learned the hard way… avoid AppFlow like that kung flu lol

Yeah we don’t have it enabled right now… Is that the general opinion across the board?

I can share an opinion on AppFlow, but the language filters would block it

Whats the profile mgmt solution and VDA version? I ad similar issues a few weeks back with an FSLogix update...

Have a look at one of the disconnected sessions from the VDA and see what is still running

Odds-on it is antivirus software or App-V client

Im sure there is a bug in that verison for this issue and you need CU2

Logoffchecksysmodule is not configured on the server

Will need to be, if a process is hanging on after logoff....

<https://discussions.citrix.com/topic/414703-disconnected-sessions-not-logging-off/page/5/>

i saw that in the forum but i didn't see anything official from citrix.

how to i determine what process is hanging on

procmon? Maybe review the CU2 release notes

like i said when i look at a server and see the disconnected states it doesn't match the disconnected session through the command line

i could have sworn there was a way to check through the command line

i checked CU2 release notes and there is some disconnect issues addressed but its not the same issue i am having.

I had this happen for a client last year, the related ADMX Citrix policies on the client's domain were out of date, after updating them, the issue was resolved. Try that to start? Simple fixes with Citrix are rare these days, but a blessing when you find them

Its per user doesn’t mater what type of session 

Not that I've seen...  would love to hear otherwise

This doesn't seem to happen in DaaS Web Studio I noticed last week. On prem Studio the column reset kills me every time.

Last I looked links to it are dead.  I have an old copy but the switches it generates are not valid for 2203+ due to changes in the installer

You can find most of the switches here, with examples as well.  Hope that helps
<https://docs.citrix.com/en-us/citrix-daas/install-configure/install-vdas/install-command.html>

tool is no longer maintained by citrix who has no plans to put the code in github

I have an old copy but the switches have changed a bit in recent releases

someone at citrix needs to pick ownership of this tool back up, going to ping the guys internally about it or ask them to release to github so community can grab it - will report back. Deff a handy little tool

What version is the existing VDA? What all Citrix services do you see?

i don't have access to the environment, i have to check...but i've seen issues come up even when those services are there but PVD is still on the system.

so on my test machine i got an error for appdisk vda pluging and i never installed it.   so i added the remove command and then the installation worked after i rebooted.

I had the same error with VDA 2212 and disabled the Citrix Graphics Tool for the moment

Its not on every sign in and I have informed the users to ignore it, so I'll leave it alone for the moment. Was planning an upgrade to 2212 to see if it was resolved, but no stress for that reason then. :slightly_smiling_face:

Because they aren’t admin workstations?

Because you can use windows tools to create a capture:
<https://msandbu.org/network-packet-trace-with-netsh-and-analysis-with-wireshark/>

Question really boils down to device registration on shared non persistent machines.  

I am pretty sure that if you AAD hybrid-join the device, you shouldn't get that prompt.

this? <https://office365itpros.com/2022/02/14/azure-ad-keep-me-signed-in-kmsi/>
I think you need to turn it on or off at a tenant level but there’s another option in there for conditional access but it applies to all apps