mslab 24.04 Hydration of DC fails after first reboot when Parent OS and VM (mslab DC) OS is Windows Server 2025

[Karl-WE]

The entire issue is described here
https://techcommunity.microsoft.com/t5/windows-server-insiders/b26085-windows-server-2025-domain-controller-fail-to-deploy-via/m-p/4119469

Scenario:
Host OS (Hyper-V): b26085 Windows Server 2025
Guest OS (DC): b26085 Windows Server 2025

Issue:
Application of DSC fails after the first reboot while hydrating the DC VM

Repro

• Install WS 2025 as host OS
• Enable Hyper-V
• Deploy mslab 24.04
• Use WS 2025 ISO during 2_createparentdisks.ps1

Workaround:
Do not use WS 2025 as a guest VM OS at the moment, for CreateParentDisks (DC) until solved.
Using WS 2025 for usual VMs, such as Management, WAC etc are not affected.

[Karl-WE]

This issue is under Investigation by Microsoft. Thank you @michbern-ms.

[michbern-ms]

@jaromirk I put together a new 'NoDCHydrate' feature which enables one to work around this for now:
https://github.com/michbern-ms/MSLab/tree/user/michbern/dchydrate

Take a look and let me know if you want me to set up a PR.

The investigation on the OS side continues, but I wanted to offer the feature as a workaround.

[jaromirk]

Oh, that's sweet! @machv what do you think?

[Karl-WE]

related #596 #589

[Karl-WE]

Today, tested with Windows_InsiderPreview_Server_vNext_en-us_26296.iso. No change in behaviour. Waiting for fix on OS level.

[michbern-ms]

@Karl-WE That's unexpected. Version 26296 has the two fixes necessary to resolve this scenario. I ran the full scenario this morning with 26296, downloaded from the Preview center:
https://www.microsoft.com/en-us/software-download/windowsinsiderpreviewserver?msockid=1818fad5388d6a5b3e2ceef839376bbb

And it worked well. I did not need to use the 'NoDehydrateDC' feature - it all just worked.

@jaromirk , @machv - would it be possible for one of you to test with 26296 and see if you can reproduce what Karl is seeing?

[Karl-WE]

Hi @michbern-ms
Thanks for sharing this information.

Hyper-V Host OS was the same build as the DC VM OS version.

I will try this again on Windows 11 24H2 as Hyper-V host OS.

[michbern-ms]

Very strange. @jaromirk @machv Would love more data points if you can spare the time.

[Karl-WE]

It's killing me softly as I thought I could finally come back to a new lab and have fun.

I am eager to help Michael. As always.

[Karl-WE]

This issue is remediated with b26311. See also #607, #604

I am closing this with a workaroud remark:

Note

Users that seek to evaluate mslab based on Windows Server 2025 please consider the following instructions:

1. MSLab Dev build, as long there is no later MSLab version release newer than 24.06.1.

2. During Step 2, (2_Createparentdisk.ps1), please use the Windows Server 2025 release ISO (mslab 24.04 Hydration of DC fails after first reboot when Parent OS and VM (mslab DC) OS is Windows Server 2025 #588 (comment)) or later for DC hydration.

fyi: @michbern-ms

Thank you everyone for their huge efforts!

> Potentially at a later time, backports from all related code changes from b26311 to b26100.1 + LCU might happen, or an updated WS 2025 ISO will be provided. If this will become available in the future this issue will be fully resolved.
@jaromirk I would recommend linking this disclaimer to mslab and Dell GEOS to avoid "frustration" as people would not expect this and using b26100 for DC instead. If you like I can write PRs for both areas.