Post-Create WIM TS in Capture Image phase fails for Windows 10 x64 Enterprise (1709) RRS feed

  • Question

  • Hello,

    I'm having issues sysprep'ng a valid deployment from a Windows 10 x64 Enterprise build using a 1709 reference image. First, I am not running into the commonly known sysprep failure out of the gates with regards to the Windows Store auto-downloading applications and thus invalidating the provisioned packages for all users thus inhibiting sysprep from running.


    • VMware ESXI 6.5 for building the VM desired for capture
    • MDT 2013 build 8443
    • ADK for Windows 10 (does support build 1709)
    • Deploying a captured WIM via WDS (or USB) using LiteTouch x64 PE image
    • Reference image is very base minus the disabling of Cortana, Windows Consumer Experience, Spotlight, Telemetry, and Windows Store application downloads (the latter is an absolute must for aforementioned reason); I'm not deploying applications through the reference image because there are far too many so the best approach is to do a "gold" image via capturing a WIM and deploying that WIM
    • Desired capture image has several large applications installed prior to sysprep (variety of Cad, Bentley software, etc.).
    • Using the base template Sysprep and Capture without any modifications to the default task sequence and unattend.xml file
    • I have a standard client deployment configured to load this WIM in the future
    • The goal is to deploy this "golden" image down to some physical workstations


    _SMSTSORGNAME=Contoso Yarharhar
    SLSSHARE=\\x.x.x.x(it's actually a valid one)\Logs$
    TimeZoneName=Eastern Standard Time
    BackupShare=\\x.x.x.x(it's actually a valid one)\Capture$




    When I launch LiteTouch.vbs from the MDT share after my VM is in the desired state, it seems to kick off just fine (boots into PE, I select capture, and it captures the WIM to the desired MDT capture share as expected. On the reboot during the Windows 10 loading screen, I am prompted with "Windows could not finish configuring the system. To attempt to resume configuration, restart the computer."  I was able to Shift+F10 to get to a command prompt and verify that not all of the services started properly and that all local users were disabled.  Since this OS is not Windows 7, the following known KBs do not apply to me:

    1. Kernel Mode Driver Framework:
    2. User Mode Driver Framework:
    3. Specific hotfix for this

    At this point, I wanted to see if the WIM was still fine or if that particular error was going to carry through in the image deployment, so I decided to run through a deploy with that WIM configured. The OS install seemed to work fine on the deploy, but as soon as it finished I received the same error I did right after the sysprep and WIM capture process from earlier.

    The BDD logs looked clean.  The SMSTS logs were reporting several of these errors (note that these logs are from the deploy machine and not the VM's capture):

    Set a TS execution environment variable _SMSTSInstructionStackString=22	TSManager	12/26/2017 10:54:31 AM	2280 (0x08E8)
    Save the current environment block	TSManager	12/26/2017 10:54:31 AM	2280 (0x08E8)
    Failed to save environment to  (80070057)	TSManager	12/26/2017 10:54:31 AM	2280 (0x08E8)
    Failed to save the current environment block. This is usually caused by a problem with the program. Please check the Microsoft Knowledge Base to determine if this is a known issue or contact Microsoft Support Services for further assistance.
    The parameter is incorrect. (Error: 80070057; Source: Windows)	TSManager	12/26/2017 10:54:31 AM	2280 (0x08E8)
    Failed to persist execution state. Error 0x(80070057)	TSManager	12/26/2017 10:54:31 AM	2280 (0x08E8)
    Failed to save execution state and environment to local hard disk	TSManager	12/26/2017 10:54:31 AM	2280 (0x08E8)

    On the deploy, I am using the default task sequences. Also, UEFI is not the path being chosen as it's a legacy BIOS Format and Partition Disk (Disk number set to 0; Windows partition name is the primary, set for NTFS using quick format based on the OSDisk variable, which is also left to its default C: value).  The Install Operating System in the Install phase is set to "Logical drive letter stored in a variable" that matches the OSDisk variable. The drive is not encrypted or locked either.  I'm not sure why the execution state and environment fail to save to the local hard disk. Any help would be greatly appreciated, and I'll post back with the SMSTS or BDD log snippets from the VM reference image where the sysprep is done later this afternoon. Thanks in advance!

    • Edited by jtcrawfo Wednesday, December 27, 2017 8:06 PM Subject/Verb agreement
    Tuesday, December 26, 2017 6:09 PM

All replies

  • Could you pull Panther logs from the system? Issues during the specialize phase could be related to drivers, unattend.xml processing failing due to formatting issues / invalid values, and so on. Either way, setupact and setuperr logs might shed some light into what is going on...


    Vacuum Breather Blog | Wing Commander Saga | Twitter

    Note: Posts are provided "AS IS" without warranty of any kind. If posts are helpful please don't forget to rate them as "Helpful" or as "Answer".

    Wednesday, December 27, 2017 7:53 AM