World of EUC

<https://docs.citrix.com/en-us/citrix-virtual-apps-desktops/policies/reference/ica-policy-settings/time-zone-control-policy-settings.html>

By default, the time zone of the user session is used. It's not very clear, so maybe it's not by default and you need a GPO setting.

Man, I never paid attention to this until now. Thanks for the quick tip

Anybody else messed around with the Timezone part? Looks like for Desktop OS.
This setting determines whether or not the time zone setting for a Single-session OS VDA is restored to the machine’s original time zone when the user disconnects or logs off. If you enable this setting, the VDA restores the machine’s time zone to its original setting when the user disconnects or logs off. For this setting to take effect, set the *Use local time of client* to *Use client time zone*.
Then set "Restore Desktop OS time zone for session disconnect or logoff". This only applies for Deskop for this example

maybe not related at all but if you have dual screens we've found the teams client has to be on the primary monitor or content sharing/video will be a black screen

Yes I am aware of this limitation. Tested on a single monitor with 1080p. Its something else :white_frowning_face:

Had a call last week with an user complaining he cannot see screen sharing during meetings, I found that clicking on the presenter icon on the lower bar fixed it

No but regarding Teams optimization in my experience the newer the better. Just have a look at the changelog. 2009.6 and 2010 

<https://support.citrix.com/article/CTX253754|https://support.citrix.com/article/CTX253754>

As written above I tested with CR 2009.6 and 2010 

What hypervisor is in use? Server or desktop VDAs?

<https://support.citrix.com/article/CTX249953>

same error msgs observed in DDC and on VDA

VDA is not having this patch, still facing this error

I was sent this  and you may already know <https://www.mycugc.org/blogs/brandon-mitchell1/2019/09/22/not-so-fast-reconnect>

Well, I just realized its for 2019 not 2016. Sorry

Yes, i'm quite unhappy about this, as a CSP

image.png

Hi, using Jabra Headsets on IGEL Ud3 (LX 11.x) devices. Teams Redirection for webcam working well, but can't use the Jabra as Audio device.  Any idea?

No you don't. There's a two way trust, you're good.

Yeah I think I’m dealing with some messed up SAML chaos - Cheers old boy 

Be assertive,  you'll sort it :laughing::rolling_on_the_floor_laughing:

I mostly see ControlUp and of course Citrix Director

uberAgent, ControlUp, Goliath, egInnovations, Nexthink, etc.

I would say that ControlUp is a must have in most Citrix setups. It's relatively simple to set up with a lot of metrics configured out of the box

If you want to be able to customize your monitoring, uberAgent is awesome

If you want out-of-the-box, Goliath is pretty good

If you want a blend of out-of-box and customization, try ControlUp

We just started a ControlUp trial and it’s been pretty good so far. Going to see what the analytics engine looks like once we have more data populated

New eg customer if any questions\input hit me up.

+1 for <#CLN1Q8614|> too. We looked at egInnovations, but went with ControlUp.

we have EG currently, feel like it's a bit cumbersome. I like ControlUp a lot more so far in teh week I've been demoing.

Thank all for your help :slightly_smiling_face:

In a previous life yes.  PVS XenApp image

<@UL5TCJ57D> If a cat has 9 lives, how many lives does a Crill have?

Thanks we're 2 weeks in and realized that setting an update policy is crucial with CS Falcon, not just optimal...

I use the one  from here:
<http://aka.ms/teams64bitmsi>

Yes use this. The current one is stored there. but always type in the machine-based parameters for citrix worker :sweat_smile::+1:

Yeah, we have a script we use, we just swap out the MSI. Cheers all!

I have some scripts to update this stuff, maybe helpful. <https://github.com/Mohrpheus78/Evergreen|https://github.com/Mohrpheus78/Evergreen>

During a migration, I had to add several times a week new VMs in an MCS catalog. Though it could be helpful to automatize it and share it with the community. Here is my first blog post :blush:
Feel free to share it! All comments are welcome! <https://stevenlemonier.fr/provision-new-vms-in-an-mcs-catalog-with-powershell/>

I would also like to know that. Having internal mic and jabra, teams is not going to save my setting which should be the default one. Using fslogix with the recommended exclusions. 

I have the same issue with 3CX. As i understand right many applications using Device GUID to identify devices. Citrix is generating new device GUIDs with every new connection(this is the reason why all the application loosing Device Assignment and reset to default) . For 3CX i found a workaround with Powershell script. But I don‘t know if it possible to do the same with teams. 

Never seen that issue.

Is the server hybrid joined?

Which fslogix and edge version. I haven't seen that yet and have already switched large environments to Edge.

I'd have to look up the versions, I'll post on Monday to bump it, but it doesn't happen if I log in through the console of the server. In addition, while I was Googling it, I found another neat edge link: <edge://sync-internals|edge://sync-internals>

<https://discussions.citrix.com/topic/408188-ms-edge-chromium-cant-sign-in-with-seamless-published-app-since-version-81041653/>

The issue that brought me to Slack in fact. The error code in your screenshot is slightly different than the one I saw initially, but I would almost be willing to bet that would just be a slight version difference.

I just checked our install here and it's still working on version 86.0.622.58 (64-bit)

Cool, thanks. I'll take a look on Monday.

If that's the case, you could use FSLogix App Masking and redirect the C:\something folder to %USERPROFILE%\Something and provide individual user settings

To be honest not quite sure. That's an input from the customer, had no assessment yet. Probably HKLM\something will came to play as well.

Cool, missed this part of FSLogix! Would it work together with UPM?

FSLogix Redirection for the win - example here <https://james-rankin.com/articles/quickpost-using-fslogix-to-create-multiple-default-profiles-and-start-tiles-layouts/>

You could use it overlaid on UPM if you *just* used this feature, yes

Turn Profile Containers on and you'd have an issue :slightly_smiling_face:

And if you have RDS CALs, which you probably have, it's free :smile:

Registry redirection is a bit trickier <https://james-rankin.com/articles/quickpost-fslogix-registry-redirection-hklm-to-hkcu-or-how-to-turn-computer-config-gpos-into-user-config-gpos/>

Wow! <@ULL6F7DRA> <@UL5TCSW3D>, thanks for super-fast and comprehensive answer! :grin:

For registry, you can try to play with REG_LINK value (symbolic links in registry), used it to fix compatibility issues before

The Machine Installer is updated twice a month apparently. It's the same on WVD AFAIK, so bit silly of them to actually duff the contract on that basis

<@UL5TCSW3D> It’s more the missing functionality like gallery view, background blurring, breakout rooms, etc and the perception that Teams on a fat client just works much better without any compromise.  (Also a 1,000 thin clients sitting idle in empty offices is a big issue ATM). 
...but yes, I’m gutted I’ve spend 11years building a great Citrix environment. 

Yeah they're all roadmapped somewhere but fat clients get them way before RDSH or VDI. No problem with performance, it's just features. Haven't heard anything from either camp about timelines

Teams for VDI will always be a second class citizen I guess. However some improvements regarding gallery view shouldn’t be too far away. Just my two cents.

I was working with Citrix/Microsoft on some Teams bug testing  and asked a lot about gallery view.  Originally i was told it was the initial roadmap called for November but now by end of year so who knows!  When it does initially come to the VDI version sounds like it will be initially limited to a 2x2 grid only (showing the 4 most active talkers only).

<@URVNXB96C> thanks, even by the end of the year with 2x2 would be something to save my sanity. Really hope this timescale comes to fruition. :pray:

Yeah i'd be interested to know about Gallery view if anyone has any more information about when it will appear.....

Save as batch and it should work. You can run it via scheduled task with the SYSTEM account. 

REM Set variables
SET SOURCE=%~dp0
SET SOURCE=%SOURCE:~0,-1%

cd "%SOURCE%"
query session &gt;session.txt
for /f "skip=2 tokens=3," %%i in (session.txt) DO logoff %%i
del session.txt

awesome we will give that a crack - thank you

<https://webcache.googleusercontent.com/search?q=cache:VzP0pWwdQfgJ:https://www.citrix.com/blogs/2012/06/05/slow-web-interfacemmc-console-crl-explained/+&amp;cd=1&amp;hl=en&amp;ct=clnk&amp;gl=us>

I feel bad when I have to use Google Cache to find my old articles :disappointed:

Btw I’ve had this issue in my environment running 7.15 on server 2016 and opened a case with Citrix.  If you have several DDCs and if they are dispersed geographically Studio is slow to launch as it has to poll each DDC. And they dont seem to execute this in parallel. Its a terrible design but they said there was nothing to be done. 

Nothing to be done, lol. They wrote the damn thing.

I’m good with anything that isn’t MMC at this point.

Hope the ctx Cloud Web Studio will come to on Prem :pray::skin-tone-2:

XenServer with so much better with Xen Orchestra, I used to recommend it to all customers

As much as I love XenServer - PowerShell implementation is not the best one :slightly_smiling_face:

XenCenter is literally the least of my worries in terms of consoles. Studio should be the main focus, given how atrociously it performs.

The fact that studio does not let me copy paste machine names has forced me to learn powershell and do most things in a command line

it the dumbest thing that you cant select a machine name in studio

But it did force you to learn PS, which is a net positive... :slightly_smiling_face:

keep it coming - I would like to share this feedback with some of our UX designers :eyes:

UPDATE THE APP LAYERING CONSOLE FOR THE LOVE OF ALL THAT IS HOLY PLEASE

Make it so you can create a machine catalog and assign those machines to a delivery group in a single operation

make every single bit of text selectable and copyable

move all the features from director into studio too so I don’t need 2 things open

or move all the features from studio to direcotr so I don’t need the MMC

that solves the selectable text issue too

<@UKJTJD4KA> WTF is the customer still using IE? 

:slightly_smiling_face: no but that's where i asked him to configure zone assignments for testing

so that didn't make any difference sadly - no idea why studio is so damn slow to load - I will get them to log a ticket with Citrix (poor buggers)

we want to use this for a customer via published app Audio and Video. Hope this will work

Just publish Explorer and have a shortcut to Teams inside it. Bingo - published desktop but actually an application as well

<@UL5TCSW3D> Publish explorer, then have them launch Teams.  I get it the session sharing will kick in, But what else behind the reason  your doing this?

It just doesn't seem as shitty an experience as publishing Teams directly

We have loads of users who run a "quasi-desktop" from explorer and we don't get billed for desktops, just for apps

<@UL5TCSW3D> I have a customer who would like to run Teams as a published app. I am having some difficulties getting Teams to even start up as a published app. Do you have any "official" statement saying that Teams is not supported as a published app? I haven't been able to find any information about it.

No, but the fact that it runs like shit or not at all should be enough

In the teams doc they talk about running it as a seamless app. But that is it. We run it, but I am still testing it out. It works, but I will make everything runs like shiz.

I would like to try James approach, putting it in a folder and publihsing the folder then have users open it. I need to play with that.

Citrix FINALLY, yesterday, updated the advisory to reflect this.

Is this all in a VDI desktop? Policy restricted, but still passes through sounds like you have a HID device that is getting blocked. But the HDX optimizations are still passing it thouhg

Same behaviour on RDSH (W2016) - non persistent desktop and VDI (Windows 10) - persistent desktop.

Right, so it say policy restricted, but still works in the session correctl?

Yes for the USB headset, so based on your doc you wrote, I think it is normal, no?

its normal for now, but I have a case on it as well.

What isn't normal (I think) is my input device (mic) empty and how S4B works with the mic? but no other app (e.g. WebEx). Do you see something in input device? I don't see any audio device blocked in the list of device (in Workspace), same level than my USB headset.

I have spent hours and hours on the policy restricted deal. I had to open a case, because I believe its a bug. Its been open since JUL 16, 2020 1:17 PM

Yes it is confusing, showing not redirected but actually working. Tbh I don't mind too much. What I don't really get is how did I block my input device, I need to figure out this and then it will sort out my WebEx issue.

I’ve seen the policy restricted thing on a mac where it was the mac policy that was restricting it, like a mac security setting that wouldn’t let the mic be used by the workspace app

citrix support were totally useless about it too

so policy restricted can mean a lot of different things

OK thanks, you refer to the USB headset, gotcha. But even with a jack, I should see something in my input. It is like the mic audio part is not redirected even if it should based on my policy. I'm probably missing something obvious.

I don’t know that you are, honestly it porbablyis a bug, I’ve seen a bunch of inconsistent stuff that makes no sense when it comes to audio redirection too

Yes it is weird ... at least it works for S4B! :slightly_smiling_face: I need to run more tests, if I found something, I'll post it there. thanks.

Yea please do. I emailed them about my case. If I here something on a fix I'll post as well.

Scenario is a single 2016 VM from which we publish RDP links to 3rd party consultants, support etc. to hop onto application servers.   If there's another way of thinking about achieving session recording, i'm all ears.

Mandatory Steps Recorder? Poor mans solution but it might just be enough?

Yep could do something with problem step recorder. Some have scripted it.

atmgr is part of WebEx Meetings, yes. Update.exe is the process that calls Teams.exe in the default shortcut, is it not?

I am not sure, but it makes sense as teams was running on the user session. Would it be safe to kill the teams update process ?

I guess there's an OOB patch for it:

<https://twitter.com/mrgrayaz/status/1329178735493353472|https://twitter.com/mrgrayaz/status/1329178735493353472>

Got hit by this too.  Pretty painful when your users lose their TGT and profile is on FSLogix.

Do you know since which version? I have 2007
Also in the Free version? I don't see it oO

It's in 2007, but I'm not sure if it's part of the free/community edition. <@UNBRPVD4G> can confirm.

Or alternatively <@UNBRPVD4G> and <@UMVNUT9C6> 

Ironically enough, my license expired today so I can't see it, haha.

It's in the 2007 release for sure, but only for the Pro Edition :slightly_smiling_face:

CalculateMaxMTU.ps1

That's basically what I did before this to figure out the MTU, Martin. Nice.

I'm pretty sure, our Firewall is the problem

I have to disable EDT on every single Site, where we have a pfSense TCP/UDP NAT in place

If I connect through the Gateway VIP, without pfSense, I don't have problems.

i use bginfo with a .blg dennis mohrmann created its really good as it shows if udp or tcp is used as well as the clients mtu size

Yes I know, it's very good. But as long as pfSense kills my EDT traffic, all will be on TCP.

you tried mtu discovery there is a reg key you can set on latest vda's and as long as clients are on latest workspace. I also had issues with edt but the mtu discovery seemed to fix it

What were your issues?
Mine are terrible latency fluctuation.
You can see windows move several seconds later. And I goes fast and slow every other minute.

We had the same issues with EDT previously, and a NetScaler upgrade to 12.1-56.22 resolved them. We upgraded QA to 12.1-58.15 and I started seeing the same issues again, so we have yet to turn it on in Prod.

I tested with NS12.1 <http://56.22.nc|56.22.nc>

with EDT its packet fragmentation that causes issues. We had connection error's / freezes was really intermittent and only affected certain users all other users on the same VDA's were fine. We have users connecting from home on adsl,3G,4G and all use different ISP's and come from different countries and some even use the corporate vpn so with the differing MTU size was killing us so mtu discovery has definitely improved things. i also did upgrade to latest 12.1 build on the netscaler as well as upgrade vda from 7.15 to 1912cu1

I have never "connection error's / freezes", only latency.

VDA 1912.0
Every Workspace app 1912 and later
NS 12.1 56.22
Always high latency

I think I've seen somebody else here claim to have the same issues. <@ULLH5JGFR>, perhaps?

I'm THANKFULL for every input. I research this issue since early 2019 I think.

We had a great VC2GO session last week with a presentation from <@ULLP2RGQ0>! Maybe helpful. Enable MTU discovery and edit the default.ica to a low value. Clients that support MTU discovery will adjust the value themselves.

Hah, I had no idea! Thanks <@ULLP2RGQ0>, also great presentation topic for 2020

on the script you just run the  .\CalculateMaxMTU.ps1 -IP adress ?

Yeah, if I remember correctly. Wrote it for a quick test about 5 years ago, so don't remember exactly :slightly_smiling_face:

<@UMB4XQE79> how could I convert the language?

This is also great from <@UL79UNCRH>
<https://github.com/endoleg/endoleg-scripts/blob/master/MTU-scan.ps1>

FWIW, we used this for the client side: <https://support.citrix.com/article/CTX231821>

There's a link in there for MTU discovery on the VDA, just a reg key.

<@ULLH5JGFR> german is an easy language to learn, danish on the other hand is a bit more difficult :D

It's not MTU. I tried everything.
default.ica value, MTU discovery.
I know the positive resaults from Sasha, but they didn't help me.

Too bad, but MTU is always a funny topic :see_no_evil:

I tested MTUDiscovery=1 the day it was available, didn't help us.

I took a look at the script from Thorsten, but it doesn't tell my why everything is laggy, even when MTUDiscovery is in place, does it?

Laggy is alwasy so hard to find. Even with Control up, Things will look perfect. Then I will see ADM show tons of Client retries.

I had a HUGE case with citrix support. We transfered several huge TCPDUMPS. They couldn't see anything.

So I upgrade my Test ENV (still behind a pfSense NAT) to:
• Win Srv 2016
• CVAD 1912 CU1
• MTUDiscovery=1
• NetScaler ADC VPX 13 67.39
• Workspace app 2010
Still laggy as hell

What CTX polices are you running? Just curious

Graphical? Isn't most of it visible in the RDAnalyzer Screenshot? What else are you mssing?

Time to isolate things. This is only external connections right. You said behind a pfSense NAT. So I assume external?

It only happens external via Gateway VPX / ICA only mode with EDT enabled.
External TCP -&gt; all good
Internal (no ica proxy) EDP -&gt; all good

I pretty isolated it to pfsense already, but they said I should fix MTU xD

Ok, So when you disable EDT and go back to TCP. Things are smoother?

If I am asking the same things you already posted, I am sorry

for the Gateway VPX what hypervisor is it?

ok, so you know its EDT then. Man that blows. Support call to the firewall vendor?

What's the MTU size of pFsense firewalls? If you lower MTU to 800 for example, doesn't that work?

Do you see packets are being send in the correct order? Because that's also something you can look at

You read my TCPdumps a few months ago, you remember?

You asked me the same, and took a look at the dumps for yourself

My memory is bad these days, to much going on

My boss had an idea about a "hardware" problem, I'll see the results tomorrow

<@U01067Q2QUV> ^^ Maybe you could also help out?

Wow, this is long thread... :slightly_smiling_face: Could someone summarize the issue/question?

The „hardware“ problem was not the issue btw

<@U01067Q2QUV>
Since ~2 years I can't use Citrix EDT. It's extremely laggy every few minutes (5.000-10.000ms) at every customer I have.
It's only laggy external via Gateway. Client -&gt; pfSense NAT -&gt; Gateway VPX ICA Proxy
No problems with TCP.
It's not MTU. MTUDiscovery=1 / current ADC Firmware / Current Workspace app.
That's a short summary.

Thanks <@ULHA2J7RT>. I realize you said it's not MTU, but right now it sounds like fragmentation-related issues due to wrong MTU. You may have covered this earlier in the thread, but to save some time:
1. What version of the VDA?
2. What OS on VDA machine?
3. What version and platform of Workspace app have you tried recently?
4. On Win client with "Current Workspace app", are you seeing a custom MTU being set by MTU discovery?
5. Have you tried using a hardcoded MTU in the default.ica?
6. What's the ADC firmware build?
7. Are you doing packet inspection at the firewall?
I'll pass along the info to the engineering team and see what they suggest we look at.

He posted
VDA 1912.0
Every Workspace app 1912 and later
NS 12.1 56.22

1. 1912 CU1 since yesterday 
2. 2016
3. Currently 2010
4. 1492
5. Yes
6. Latest 13.0 VPX on vmware since yesterday
7. No
I had a HUGE netscaler case over months and they found nothing after several triple tcp dumps (client/netscaler/VDA simultaneously)
They ultimately said it must be a pfsense issue. You want the case number?

Thanks! And yes, please share the case number
Also, anything special about the pfsense config? Or just doing basic firewall allow/block based on ip:port and protocol?

Will send the number later. 
No, not afaik. As you described it. 

Yeah, that's pretty standard, though.
Cool, thanks for the info. I'll pass this along and circle back with you.

I sent to you both case numbers in a private message

I found something new today, I obviously missed until today. :disappointed_relieved:
It seems like only mouse/keyboard input are laggy, but not the screen. I noticed that, as the minutes skipped on the systray clock, while nothing happend, cause it was laggy.
Afterwards all my mouse and keyboard inputs were executed after a delay.

I than started to let a visible stop watch run in my session, and during heavy lags, which are like 10 seconds+ the watch keeps running, but the EDT session doesn't honor any input, but catches up after the lag is gone.

<@ULLH5JGFR> <@UNBRPVD4G>
Do you know something about my new findings, that only input mouse and keyboard seem to be affected, while the visible screen still works?

It depends in my experiences. This could be because of multiple (DPI mismatch)  monitors, Visual Effects, Cursor Shadow(mouse- Uncheck Enable Pointer Shadow, and Uncheck allow them to change mouse Pointer. ) As I recall you are using the High Server Scalabity Ctx Policy?

No. At least not that I’m aware of. If I remember correctly this is off by default. 

Ray, I try to record a video demo tomorrow, because I believe none of you really knows what I’m experiencing (no offense), because it’s really hard to describe for me. 

Yeah a video says more than a 1000 words :wink:

<@UNBRPVD4G> <@ULLH5JGFR> <@U01067Q2QUV> Here is a video demo with annotations, I hope this makes clear what we are seeing:
00:17 Input Lag start
00:26 Lag catches up
00:38 RDAnalyzer shows the lag
... Normal
01:17 Next Input Lag start
01:20 small catch up
01:24 Next Input lag
01:33 input color change, which might get unnoticed due to lag
01:38 changed color shows on screen
01:52 RDAnalyzer shows current input lag spike
<https://files.meinekleinefarm.net/f/bc0588eccc6e454bb449/>

Wow that's bad. I haven't seen anything that bad unless something chewing up the Nic resource. Man, it's time for a Citrix support call with you firewall vendor. IMHO

I know it's painfully. I wish I could give u advice to help fix it.

We have a support call with the Firewall vendor since sunday

They are working on it, and made suggestions, we might implement next saturday

Noo as I can see for 60% of users. I did few checks. Probably it’s not calculated when there is no traffic from users. 

It works with HDX but some users reporting N/A. Probably because of no network activity. Any idea ho to enable calculations for idle connections?

there is a citrix policy to enable ica round trip for idle connections

I need for “network latency” - ControlUp do not support Citrix RTT :confounded: