Finding in the Wild | Citrix MCS – The VDA Starts, but the Domain Trust Is Broken

An issue that can be particularly disruptive in non-persistent Citrix environments. The VDA appears to start normally after a reboot. The VM is reachable and the network connection is up, but users can no longer log on. ❌

An issue that can be particularly disruptive in non-persistent Citrix environments.

The VDA appears to start normally after a reboot. The VM is reachable and the network connection is up, but users can no longer log on.

❌ Windows reports a broken domain trust relationship
❌ User logons through ICA or RDP fail
❌ A restart does not permanently resolve the issue
❌ In the worst case, multiple machines in an MCS catalog are affected

🔍 What is the actual cause?

With Citrix Machine Creation Services (MCS), each machine receives its own identity.

This is supplied through components including the MCS Identity Disk.

It contains the machine-specific identity information required for the correct association with the Active Directory domain.

If the machine account password or trust information no longer matches the AD computer account, the domain trust relationship may fail.

Possible triggers include snapshot rollbacks, incorrect identity information or manual changes to AD computer accounts.

🛠️ My troubleshooting approach

First, I check the secure channel on the affected VDA:

Next, I check:

✔️ Netlogon events
✔️ The associated AD computer account
✔️ The MCS Identity Disk assignment
✔️ Machine status in the Citrix Broker

And here is the interesting part:

Citrix provides Repair-AcctIdentity for targeted repair of the machine identity.

It can repair the machine password and trust information without immediately rebuilding the entire machine catalog.

Example:

Important: the affected VM must be powered off and in maintenance mode.

💡 My takeaway

For a broken secure channel in a non-persistent MCS environment, I would not simply remove the computer from the domain and join it again.

MCS manages the machine identity. A manual repair inside the VDA may therefore be lost at the next restart.

Check the MCS identity first, then perform a targeted repair instead of immediately rebuilding the entire catalog.

Finding in the Wild | Citrix Workspace App 2607 LTSR x64 – StoreFront SSO Stops Working

A Citrix Workspace App upgrade, and suddenly automatic sign-in to StoreFront stops working. These issues are particularly interesting because the cause is not necessarily where you first expect it. The symptoms: ❌ Windows logon works normally ❌ StoreFront

A Citrix Workspace App upgrade, and suddenly automatic sign-in to StoreFront stops working.

These issues are particularly interesting because the cause is not necessarily where you first expect it.

The symptoms:

❌ Windows logon works normally
❌ StoreFront requests a username and password again
❌ Domain pass-through authentication no longer works
✅ SSO worked before switching to the native x64 Workspace App

The first suspects are often Kerberos, GPOs, StoreFront or the SSON component.

The actual cause:

Citrix documents a known issue with domain pass-through authentication in Workspace App 2607 LTSR x64.

Registration of the cck.dll component may be missing or incorrect in the 64-bit registry branch.

🛠️ The workaround

On the affected client, registration can be corrected using the following command:

Then exit Workspace App completely and test the sign-in again.

My takeaway:

If StoreFront SSO suddenly fails after a Workspace App upgrade, I would not immediately dismantle the entire authentication chain.

First, I check the client version, architecture and registration of the components involved.

A small registry entry can make the decisive difference here.

⚠️ This workaround applies to the documented symptoms and should be tested before a broad rollout.

Finding in the Wild | Citrix UPM – First Logon Fails, Second Logon Works

This week, I encountered an interesting issue with Citrix Profile Management (UPM) at a customer. After restarting the VDAs, various startup applications did not run during the first user logon. The symptoms: ❌ Applications from the Windows Run

This week, I encountered an interesting issue with Citrix Profile Management (UPM) at a customer.

After restarting the VDAs, various startup applications did not run during the first user logon.

The symptoms:

❌ Applications from the Windows Run keys did not start
❌ Shortcuts in the Startup folder were ignored
✅ Everything worked on the second logon
❌ The issue returned after a VDA restart

Initially, I focused on the usual suspects: GPOs, WEM, logon scripts and Windows Explorer.

But the actual cause was in the user profile.

🔍 What did I find?

During the investigation, the following registry key caught my attention:

It contains session-related information for Windows Explorer.

In the affected environment, Citrix UPM persisted this key in the user profile and restored it at the next logon.

This apparently carried over outdated session information that affected initialization of the startup applications.

🛠️ My solution

I excluded the registry key from profile synchronization in the Citrix UPM configuration.

Citrix UPM GPO:

Profile Management → Registry → Exclusion list

Added the following entry:

Software\Microsoft\Windows\CurrentVersion\Explorer\SessionInfo

Next, I cleaned up the SessionInfo data already stored in the affected test profile and tested logon again.

The result:

✅ Startup applications ran again on the first logon
✅ The issue did not return after a VDA restart
✅ No changes to the applications themselves were required

💡 My takeaway

Not every logon issue is a problem with GPOs, WEM or Windows itself.

Especially in non-persistent Citrix environments, it is important to consider which registry information actually needs to persist between sessions.

In my view, session-related Explorer data does not belong in a permanently synchronized user profile.

I will therefore consider excluding Explorer\SessionInfo in my future UPM baselines, after assessing each environment.

⚠️ This exclusion is not a universally confirmed Citrix fix and should be tested before a production rollout.

Finding in the Wild | Citrix VDA – Microsoft Edge Suddenly Loses Internet Access

This week, while investigating a Citrix issue, I came across some interesting behavior. The starting point: Windows 11 with Citrix VDA 2507 LTSR CU2 in a Remote PC Access scenario. After installing the VDA, internet access through Microsoft

This week, while investigating a Citrix issue, I came across some interesting behavior.

The starting point:

Windows 11 with Citrix VDA 2507 LTSR CU2 in a Remote PC Access scenario.

After installing the VDA, internet access through Microsoft Edge became unreliable.

The curious part:

✅ Network adapter active
✅ Windows Firewall was not the cause
❌ Edge could no longer reliably reach external websites
❌ Downgrading the VDA did not help

Access worked again only after completely uninstalling the Citrix VDA.

🔍 Where was the cause?

Not in the conventional network stack, but in a Citrix HDX hook:

CtxWebAuthnHook.dll

This hook is part of Citrix WebAuthn integration and supports the corresponding authentication scenarios.

The affected environment also used a proxy auto-configuration (PAC) file.

This particular combination apparently caused a conflict between the Citrix hook and Microsoft Edge.

🛠️ What was the solution?

The decisive clue was the process-specific hook configuration for:

CtxWebAuthnHook\msedge.exe

After removing the corresponding registry entry, internet access worked again.

The VDA could remain installed.

⚠️ Important: this is a targeted workaround for the affected configuration. Before changing Citrix hooks, assess potential effects on WebAuthn and FIDO2 authentication.

💡 My learning from this finding:

If browser issues suddenly appear after a VDA installation, do not limit your investigation to DNS, firewall, proxy or the network adapter.

The Citrix API hooking configuration can also affect individual applications.

In environments using PAC files, WebAuthn and FIDO2 in particular, it is worth examining the loaded Citrix hooks.

A working network stack does not automatically mean the browser can communicate without issues.

Finding in the Wild | Citrix 2607 LTSR – Upgrade Succeeds, but Software Licenses Suddenly Become Invalid

This week, I upgraded a customer’s Citrix environment to the current CVAD 2607 LTSR release. Initially, the upgrade went completely smoothly. ✅ VDAs started correctly ✅ Registration with the Delivery Controller succeeded ✅ Applications could be launched ✅

This week, I upgraded a customer’s Citrix environment to the current CVAD 2607 LTSR release.

Initially, the upgrade went completely smoothly.

✅ VDAs started correctly
✅ Registration with the Delivery Controller succeeded
✅ Applications could be launched
✅ No anomalies in Citrix Director

An apparently successful upgrade.

But shortly afterwards, individual business applications suddenly reported licensing issues.

❌ Invalid license
❌ Hardware ID changed
❌ Reactivation required

Initially, everything pointed to an issue with the licensing software.

However, the actual cause was a change in Citrix MCS.

🔍 What did I find?

Since VDA 2511, Citrix ensures that MCS-provisioned Windows machines have a unique MachineGuid.

Previously, multiple MCS machines could inherit the same value from the master image.

With newer VDA versions, including 2607 LTSR, this machine identity is assigned uniquely and persists across restarts and image updates.

A sensible technical improvement.

However, some business applications use the Windows MachineGuid as part of their licensing fingerprint.

If that value changes during an upgrade, the software may identify a new machine.

🛠️ How did I check this?

The MachineGuid can be read from the registry:

Comparing it with the previous machine identities revealed the change after the upgrade.

This also explained why the affected applications required license reactivation.

What am I taking into future upgrades?

✔ Document MachineGuid before the upgrade
✔ Check licensing dependencies of business applications
✔ Test the new VDA version on pilot machines first
✔ Validate license activation after the upgrade
✔ Coordinate reactivation with the vendor if needed

⚠️ Simply restoring the old MachineGuid or copying it to multiple VDAs is not a sound solution.

💡 My takeaway

The Citrix upgrade was technically successful. Nevertheless, a background change directly affected users’ applications.

Not every issue after a Citrix upgrade is a classic Citrix fault. Sometimes a technical foundation changes that other applications depend on.

Citrix documents this behavior under PMCS-58288.