After installing the September 2026 cumulative updates—KB5122871 on Windows Server 2025 (build 26100.33438) and KB5122882 on Windows Server 2022 (build 20348.5622)—every host in our mixed RDS collection began experiencing the same severe session-management hangs. The hosts are VMware VMs using RD Connection Broker, MFA at logon, and third-party endpoint protection. They were stable with the August updates.
The issue usually begins after a session disconnect or reconnect fails to complete. New RDP connections remain stuck at "Connecting…" and never reach the logon screen, with TerminalServices-RemoteConnectionManager/Admin event 20498 reporting that the client connection took too long. Existing users cannot log off or disconnect cleanly. Task Manager and other Windows applications, especially Settings-related components, also hang when they call into the Local Session Manager. Winlogon event 6005 reports that SessionEnv is taking too long to handle a disconnect notification.
A normal restart eventually hangs as well, so the host has to be hard-reset. The problem returns later the same day. Both updates mention an improvement to Remote Desktop audio redirection, but neither currently lists an RDS-related known issue.
We are testing removal of the September cumulative update from part of the collection while leaving one updated host as a control. Has anyone seen similar behavior after KB5122871 or KB5122882, confirmed an FSLogix-related cause, or found that rolling back the LCU resolved it?
3 Answers
There is no official RDS issue listed for either update yet, so this may be ahead of Microsoft’s documentation. Also check the update packaging when testing a rollback: these updates include a servicing stack update and an LCU. Removing the KB by number with wusa may fail. Use DISM to list the installed packages, identify the LCU package, and remove that package directly. The SSU will remain installed, which is expected.
While testing, collect the FSLogix profile logs from a host while it is hung. A profile container that never finishes attaching can wedge one session and make later Local Session Manager operations appear to hang, which can look very similar to an operating-system update regression.
Before treating the cumulative updates as the only cause, compare the affected hosts’ FSLogix versions and profile-container logs. A failed or indefinitely pending profile attach can create the same pattern: one session gets stuck, then operations involving session state queue behind it. A clean rollback result across otherwise identical hosts would make the update connection much stronger.
I saw comparable behavior on Server 2016 RDS hosts after removing KB5123099, and the issue stopped after the reboot. Since that update appears to contain a similar Remote Desktop audio-redirection change, testing without KB5122871 and KB5122882 seems worthwhile. Keep the updated control host isolated enough to compare connection, disconnect, and restart behavior.
Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures