Appstacks not mounting - for some users on a regular basis

[Lansti]

Hi.

I see from time to time that some of our users who logs on our Horizon 8 desktops, that the appstacks do not mount.
I will say that we see a couple of users almost every day.
Quick fix for this is to log off windows, and log on again.

I'm troubleshooting this to firgure out why, so I can fix this issue/problem.

We are running Horizon 8 non-presistent desktops with Appvolume version 2212.

I checked the svservice client logs located in %programfiles(x86)%\CloudVolumes\Agent\Logs, and see that virtualization is disabled.
[2024-09-02 05:20:56.137 UTC] [svservice:P2024:T7316] Message: "Error from Manager "appvoltarget.domain.com" (error code 400):
Unable to login because Computer "Computer <domain/Computername$>" is offline, last seen online at "2024-08-29 21:24:20 UTC"
Virtualization is disabled." (hToken 0000000000000790)

From my appvolume instance i see that i have two entries:


Where one of them are "marked" as conflicted, and i just assume that this is the issue I've to deal with.

But after a futher look into the appvolume target - \directory\computers i can see that I have several desktops with the same "Conflicted" marking...


Its seems to me that i have a bit of a cleanup job to do, but not sure what to do to get rid of these conflicted clients.
Any suggestions..?

I also need to have a new look into my configurations, I can see that i have tons of these messages in my "System messages":
LDAP bind failedon "dc.domain.com ", error "49": 8009030C: LdapErr: DSID-0C0905E4, comment: AcceptSecurityContext error, data 52e, v3839

I have tried to remove and add the server in the App Volumes Manager Servers, updated the Machine managers with correct user/pass, so it seems to be working, but still i get these messages.

My AD Domain registration looks like this:

 

Hope some wise heads here can point me in the right direction.

Thanks

[Lansti]

I've done some further troubleshooting, and see that clients marked as conflicted also appear to be online, and the other client is offline, and I think this is the problem we are seeing here.
That Appvolume tries to mount the disks to a machine that is offline.

When I look at the intervals for scheduled jobs, "Refresh Computers" says 2 hours and 11 minutes.
"Refresh Machines" says 22 hours.
It seems that I need to adjust these intervals.

In the documentation:
Refresh Computers
Ensures that the state of each computer is online.
Refresh Machines
Ensures that each virtual machine exists. Removes the virtual machines if they no longer exist
and are not in use.

Hope that adjusting the times for scheduled jobs will contribute to more efficient cleaning of the database so we do not get too many entries that say "conflicted"

[Philipp Siebers]

Hi Lansti,

Yes, the computer status is directly involved when a package/AppStack gets attached. An offline computer won't get any attachments.

In addition to adjusting the intervals, you could alternatively try disabling the agent session cookie (Configuration -> Settings -> Advanced settings).
In one of my deployments, this improved detection of whether a computer was online or offline.

 

Edited September 2 by Philipp Siebers

[Lansti]

Thanks Philipp, I'll have a look at disabling the session cookie, just need to see if my change of intervals of the scheduled jobs have an impact first.