CloudCache issue after relogon

[Dmitry_Malyshev]

Hello everyone.

Trying how to figure out the following issue.

Our configuration:
FsLogix Apps 2.9.8228.50276
FsLogix Cloud Cache feature enable and configured using 2 SMB file shares on different File servers

Windows Server 2016 OS version
VDA Citrix LTSR 1912 CU6
Citrix farm 1912 CU6 with many servers
Infrastructire is on AWS, t3xlarge instance type, EBS volume with GP3 type.

No problem with limitations from AWS side found

When user is closing hosted application and trying to open another one in short period of time (several seconds) user will have starting session on different server because of load balancing and get profile which is not under FSLogix control and use just local profile from the scratch.

Logoff process itself on Application Server takes approximently 5-7 seconds, but merging local cache from AS to File Share takes more time myabe 15 seconds. That's why FsLogix is unable to handle new session on another server correctly. In this case we are getting this literraly logic error in ErrorLog

Error attempting to acquire shared reader lock. File: 'D:\ProgramData\FSLogix\Proxy\UserName\Profile_UserName.VHDX', Owning MachineId: 'servername.fqdn'
Trying to acquire lock with machine id 'NewServerName.FQDN', but lock has already been acquired with a different machine id 'PreviousServerName.FQDN'

I was trying to playing around with LockedRetryCount and LockedRetryInterval settings but without success.

I totally understand that logoff process takes some time, but I want to adjust some configuration for waiting when .lock file on the share will be released and FsLogix will be able to use user's VHDX as usual. So yes, logon duration will be impacted but not a big problem for sure as the topic one.

Please share your experience using CloudCache configuretion

[#2]

A few things you may want to check out:
-Update to the latest version of FSlogix as there's a few bugs that have been fixed
-You may want to turn on the GPOs "Prevent Login with Failure" and "Prevent Login with Temp Profile", while this may cause errors anway, at least your users won't log in with a local profile
-We are using VMware Horizon with Instant Clones and Cloud Cache configuration. The new FSLogix compaction feature is on by default and doesn't work well with Cloud Cache. It works well with a single server setup. You may want to turn it off until Microsoft properly supports Cloud Cache with VHDX compaction. You can read here that there's a possibility of "poor user experience" https://learn.microsoft.com/en-us/fslogix/troubleshooting-vhd-disk-compaction#how-does-disk-compacti...
In our case it simply doesn't work. In your case, it may significantly reduce the sign-out time and the .LOCK release.

[#3]

Thank you for reply.
We use not the latest fslogix version because we faced a lot of issues during testing in Dev environment, but about month ago MS released a hotfix with a lot of bugfixes. And yes, we must
turn off VHDX compaction in our case.
We will test the latest version and test your recomendations.
Btw, do you what the following does means?
for example this GPO setting "Locked VHD retry count" has description "Supported on FsLogix Profiles 2.5" but I see in the log during logon process that "Profile format version 2"
What is the Profile version and how to update or control it?
Didn't find this in google.
Thank you

[#4]

I could be wrong here but I think VHD is Profile format version 1 and VHDX Profile format version 2. As long as your FSLogix version is at or over 2.5, Locked VHD retry count will apply.

[#5]

I have tested the latest version FSLogix 2210 hotfix 1 (2.9.8440.42104) and seems like they fixed the behavior.
Currently Flogix is trying to attach VHDX even if it Locked by another machine according to these settings LockedRetryCount and LockedRetryInterval.
In this case it is suitable for me so far.
Thank you

Log file for example
[07:42:43.283][tid:000011f0.00003bf0][ERROR:000000a7] Operation 'AcquireExclusiveLock' failed. Retrying 0 time(s) at 0 second intervals (Unable to lock a region of a file.)
[07:42:43.287][tid:000011f0.00003bf0][INFO] Configuration Read (DWORD): SOFTWARE\FSLogix\Profiles\LockedRetryCount. Data: 7
[07:42:43.287][tid:000011f0.00003bf0][INFO] Configuration Read (DWORD): SOFTWARE\FSLogix\Profiles\LockedRetryInterval. Data: 3
[07:42:43.320][tid:000011f0.00003bf0][ERROR:000000a7] Operation 'AcquireSharedReaderLock' failed. Retrying 7 time(s) at 3 second intervals (Unable to lock a region of a file.)
[07:42:46.321][tid:000011f0.00003bf0][INFO] Retrying operation 'AcquireSharedReaderLock' 1/7
[07:42:49.356][tid:000011f0.00003bf0][INFO] Retrying operation 'AcquireSharedReaderLock' 2/7
[07:42:49.381][tid:000011f0.00003bf0][INFO] Operation 'AcquireSharedReaderLock' was successful
[07:42:49.386][tid:000011f0.00003bf0][INFO] DiskContext::lockWithStrategy returning after 6172 milliseconds
[07:42:49.591][tid:000011f0.00003bf0][INFO] Profile VHD Path: D:\ProgramData\FSLogix\Proxy\userName\Profile_UserName.VHDX