AppReadiness Service and Black Screen

Brass Contributor

The whole Appreadiness service and black screen issue has been floating around for a couple of years now, but it's now raised its head again in WVD, at least where we're using FSLogix.

 

We have two WVD setups, one with FSLogix and one with no profile solution.

 

On the FSLogix one we're consistently getting black screens at login (although Ctrl+Alt+End works and you can run, say, notepad from Task Manager) but it eventually comes to life after five minutes.

 

In the event viewer the following error is associated with each black screen:

"A timeout (30000 milliseconds) was reached while waiting for a transaction response from the AppReadiness service."

 

Standard multi-user Windows 10 enterprise from the markteplace with is 1903 (although Windows itself is offering 1909 as an upgrade which I'm guessing is WVD supported, but I've not seen anything confirming this).

 

Anyone else seen this or got a solution?

55 Replies

I've dug into this a little more.
During the black screen, if I kill the App Readiness service then the black screen instantly disappears.
Given that App Readiness should only run during the first log on of a new user, I think that it's kicking in before the FSLogix profile has fully mounted, i.e. :
Windows 10 thinks it's a new user
AppReadiness triggers
Profile container fully mounts
Appreadiness clashes with the now mounted profile.

Have you found any fixes for this. I'm running 1909 and have the same issues at the moment. If I disable AppReadiness the login process is pretty much instant. I don't understand why things can be running smoothly then all of a sudden this issue occurs.

@Greig Ritchie we are having the same exact issue it appears with black screen at login that takes 5min or so to go away before our wvd desktop loads.  Were you able to resolve?  I plan to open a ticket.

 

Thanks,

Ed

 

Hi there, we are having the same issues as well. Any clues gratefully received.

Phil

@Philworsley2   it seems our problem just disappeared.  Microsoft setup our hosts to perform a memory dump for when it happens and it has not returned!  This is what they said "Checking information we have available it appears that this issue can appear because of issues with updates or even because of FSLogix"  So maybe some updates came through and fixed our issue?  I don't know but I'm just hoping the problem fixed itself.

 

Best,

Ed

 

I think we already knew it was FSLogix so I don't know why they thought that was new information! :beaming_face_with_smiling_eyes:

The only way we managed to resolve it was with daily reboots.

@OffColour1972  Kindly check the FSLogix storage account space, kindly expand the storage (File share) to fix the issue.

@OffColour1972 What are you thinking the black screen issue has to do with FSLogix?  We are also having this happen occasionally and MS is trying to figure it out.

@OffColour1972 Using Windows 10, Version 2004, Multi-Session, WVD, FSlogix, and Azure files... all users started getting this after a few weeks use of WVD. True, if I disable app readiness service and reboot the login is quick and works as expected. However, Outlook and Office issues start arising and it has to be enabled again. All updates are installed but no fix found at present. Anyone else found a fix?

@Sponge405 We reboot our VMs daily to save cost and still have the issue. But, since we know "App Readiness" is part of the problem. I ran a performance test on a VM and rebooted. The test linked Skype to using a lot of CPU and time on loading, even though we don't use it or even login to it. I downloaded Autoruns and stopped One Drive (since we don't use or login to that either) and Skype from loading at start up and from loading browser extensions just to be sure. I then rebooted the VM and the desktop appeared after a black screen for approx. 10 seconds rather than 1-3 minutes. I am not sure why this would affect app readiness unless it skips trying to configure registry keys etc. on start up for one drive and skype. Now I need to try and implement on mass.

Okay, as close as I can get to a solution. The other attempts helped but this is the solution so far...

 

In the event viewer under administrative events you will see ERROR relating to either App Readiness or App-Model Runtime both pointing to something like... "8wekyb3d8bbwe" saying that is corrupt or install failed.

 

You will also see the profile it is attempting to apply it too such as...

S-1-5-21-527237240-2025429265-725345543-1124

 

Go to the following key and confirm the name of the user with the issue, to be fair its probably all users.

Computer\HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\ProfileList\S-1-5-21-527237240-2025429265-725345543-1124

 

Next, go to the following key

Computer\HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Appx\AppxAllUserStore

You will see entries relating to the above... "8wekyb3d8bbwe" make a backup and then delete all entries

 

Next, go to the following key

Computer\HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Appx\AppxAllUserStore\S-1-5-21-527237240-2025429265-725345543-1124 make a backup and delete the entry or all users if effecting everyone

 

You can then login and you will be on your desktop in around 10 seconds, I cant get it lower than 10 seconds at present. Everything continues to work as expected without any issues.

 

I have applied the above to a few machines now and it is the only thing that works at present!

@Sponge405 Can you say soemthing about the Outlook and Office issues you had after disabling AppReadiness? We also want to disable the service as users are unable to work while it is enabled. We've had this problem since Windows updates from July. We have noticed that the Settings button is unresponsive when AppReadiness is disabled

@MatthewHurley Basically, Outlook kept asking for the password even though it was entered and correct and wouldn't update.  Our issues maybe somehow linked to MFA since we use it throughout. The solution I found earlier only works if you script it every time you login otherwise the login creeps back up and you end up waiting again. Also, by removing the reg keys you end up removing downloaded apps from the store or causing other issues. :cry:

@Sponge405 Thanks for your reply. I already noticed on the second login that Office 365 SSO doesn't work when the App readiness service is disabled. 

We have this issue on a provisioned Windows 2019 RDP server but the symptoms are the same: blackscreens which take anything up to 15 minutes and when you do get a desktop then the start button and search are unresponsive. CRTL-ALT-DELETE works but you can't get a desktop. Disabling App readiness removes the black screen problem but leads to other issues as I mentioned before. One of the consequences of disabling App readiness is that the settings button is also unresponsive. 

It's definately related to App readiness so I'm going to play around with the registry keys and see if I can pinpoint the one causing the issue. 

BTW, we have the problem with black screens since Windows Updates of July 2020, although Office 365 also got it's half yearly update at the same time. In the meantime we're being forced to use an older image so that people can, you know, actually work. :( 

Try  changing app readiness.  also try the new rool up up date  @MatthewHurley 

left out  change it to automatic as a work around.  but they update may fix it @Steven Blatt 

@Steven Blatt Hi Steven, thanks for the tip. The machine is already fully updated and we still have the same issues. 

We are experiencing the exact same problem. Black screens after login which disappear when we kill the app readiness process. We have opened a support case with Microsoft without any luck so far.

@ausumACR I updated our gold image this week and released it to a test VM today. So far I've been able to log on to the vm 4 times without seeing a black screen. 

 

Here are the changes I made:

I installed the Windows Updates from September 2020 (KB890830, KB4570333 and KB4570332). I uninstalled the Client UI for App-V and disabled the Microsoft App-V Client service as we don't use it anymore. I also disabled the Citrix Workspace Updater service.

 

Until now we have been using Windows Defender for AntiVirus but we are currently migrating to SentinelOne. I installed the SentinelOne agent which also disables Windows Defender. We were using Sophos Endpoint protection on the file server where the FSLogix profiles are stored. Last week we removed Sophos and installed SentinelOne on the file server. Maybe this has something to do with it. As we are a 24 hour organization, users need access to the file server 24/7 so we don't manage to install Windows updates on the file server as regularly as we would like. We have to schedule a window to install updates on the file server so we took this opportunity to update and reboot the file server. 

 

Any one (or none) of things could have caused the improvement.

 

I still don't trust it so I'm not ready to release it to our users yet, but if anybody else wants to try any of the suggestions let me know how it works out. I will try later in the week or the start of next week if the improvements are still there (don't want to have the virusscanner migration muddying the waters) and then post my results