MicrosoftI installed this version hoping it would solve some of the APPX issues we are having. It became immediately apparent that logoff's are no longer 100% successful. We have only about 5 users on our test server, and 3 of them experienced a logoff issue. The log shows the following sequence of events (which don't give us a lot of info to go on):
[11:22:44.584][tid:00001120.00003d50][INFO] Configuration setting not found: SOFTWARE\FSLogix\Profiles\ControlTileService. Using default: 0
[11:22:44.611][tid:00001120.00003d50][INFO] Status set to 1: Cannot load user's profile
[11:22:44.611][tid:00001120.00003d50][INFO] ErrorCode set to 1064 - Message: An exception occurred in the service when handling the control request.
[11:22:44.611][tid:00001120.00003d50][ERROR:00000428] invalid map<K, T> key (An exception occurred in the service when handling the control request.)
[11:22:44.613][tid:00001120.00003d50][INFO] Configuration setting not found: SOFTWARE\FSLogix\Profiles\ShutdownOnUserLogoff. Using default: 0
[11:22:44.613][tid:00001120.00003d50][INFO] Configuration setting not found: SOFTWARE\FSLogix\Profiles\RebootOnUserLogoff. Using default: 0
[11:22:44.613][tid:00001120.00003d50][INFO] Configuration Read (DWORD): SOFTWARE\FSLogix\Profiles\ProfileType. Data: 0
[11:22:44.613][tid:00001120.00003d50][ERROR:00000428] UnloadProfile failed. Version: 3.25.202.4223. User: XXXXXXXX. SID: S-1-5-21-XXXXXXXXX-XXXXXXXXXX-XXXXXXXXXX-XXXX. SessionId: 20. FrxStatus: 1064 (An exception occurred in the service when handling the control request.)
[11:22:44.613][tid:00001120.00003d50][INFO] unloadProfile time: 367 milliseconds
The profile disk is left connected, and the user is unable to access the profile disk on subsequent logins because it is "in use". The situation requires and administrator to dismount the disk, which allows it to be used on the next login.
Noting that this did not happen to every test user, but it did happen to 60% of our test users.
Reverting to 2.9.8884.27471 resolved the issue for us.
MicrosoftThank you TreyContello for reporting this. Can you please provide as much detail as possible about your VDI environment, Windows OS (build), and FSLogix configuration? If you have details that you don't want publicized, please send me a DM and include your case number if you have one.
cc: Soulwaxer , NickK-Bict
FYI - Support for FSLogix does not mean direct access to the Product Team. I do respond and follow these comments, especially when tagged.
Windows Server 2019 as RDSH. FSLogix configuration:
CleanupInvalidSessions - Recently enabled due to this problem
VHDCompactDisk - Disabled
Profiles
AccessNetworkAsComputerObject - Disabled
ConcurrentUserSessions - Disabled
DeleteLocalProfileWhenVHDShouldApply - Enabled
Enabled - Enabled
GroupPolicyState - Enabled
InstallAppxPackages - Enabled
IsDynamic - Enabled
KeepLocalDir - Disabled
OutlookCachedMode - Disabled
RoamIdentity - Disabled
RoamSearch - Disabled
SizeInMBs - 0x00007530 (30000)
VHDLocations - \\domain\dfs\profiles
VolumeType - VHDX
VolumeWaitTimeMS - 0x00004e20 (20000)
This is clearly not an environment or configuration problem, there have never been any major issues with FSLogix in our environment. Reverting to the previous version immediately solves the problem. This is a bug or problem with this version. I can easily reproduce by rebuilding the server on the old FSLogix version and there are 0 issues. However, if I rebuild the server and install this new version the problem is immediate.
Support really only has the ability to review settings and ensure we are doing everything properly (verify and confirm). Since I already know we are doing everything properly, support in this case can't really help -- only the product team can solve the problem by issuing a hotfix. I understand the requirement to work through support to gather info, but once it is confirmed that the software version is the problem support can't do anything--only the product team can.
I'm glad to have an acknowledgement of the issue and grateful for any assistance. Note that several other folks have indicated the same trouble here, so it's not just me...
