none
"Updating head node configuration" failes due to "The account is not authorized to log in from this station"

    Question

  • I have recently upgraded from HPC Pack 2016 to HPC Pack 2016 Update 1, which caused some problems (https://social.microsoft.com/Forums/en-US/aa4a346a-6629-4a7c-a2e6-a088281309c1/upgrade-to-hpc-pack-2016-update-1-fails-because-sf-is-screwed?forum=windowshpcitpros). The cluster is, however, up again now, but baremetal deployment still does not work - I cannot add images. I further investigated this issue and found that the HPC service seems to be unable to configure WDS.

    In the "Operations" subtree of Resource management, I see a lot of operations named "Updating head node configuration", which have failed. This happens for all three head nodes at different times. The operations performed are

    Time Message
    23.04.2018 12:47:07 Reverted
    23.04.2018 12:47:07 Stopping service HpcDeployment on node VESTA2
    23.04.2018 12:47:07 The operation failed due to errors during execution.
    23.04.2018 12:47:07 The account is not authorized to log in from this station.
    23.04.2018 12:45:14 Locking the Windows PE image for changes.
    23.04.2018 12:45:04 Retrying operation, retry count 1
    23.04.2018 12:45:04 Starting service HpcDeployment on node VESTA2
    23.04.2018 12:45:04 The account is not authorized to log in from this station.
    23.04.2018 12:44:16 Locking the Windows PE image for changes.
    23.04.2018 12:43:25 Locking the Windows PE image for changes.
    23.04.2018 12:43:25 Updating the hosts file for the WinPE image
    23.04.2018 12:43:25 Enabling WDS and binding to: 10.35.255.254
    23.04.2018 12:43:25 Updating the local hosts file based on changes to nodes in the cluster

    A very similar error that occurs from time to time is about the DHCP configuration. As the error reported is in both cases "The account is not authorized to log in from this station.", I checked the Administrators group on all head nodes. The computer accounts of all head nodes are members of the group on all machines. The configuration of all compute nodes can be discovered without any error.

    I assume that the problem is somehow caused by the "rough upgrade process" and that some permissions are missing, but I have no idea where to look ...

    Monday, 23 April 2018 11:29 AM

All replies

  • The event log has some more information, so I now think the issue is related to dism and the WDS:

    Log Name: Microsoft-HPC-Management/Admin
    Source: Management
    Event ID: 6100
    Details:

    The operation 'Updating head node configuration' failed to run correctly. The operation was initiated by the user: SYSTEM.
    The session log for this event is as follows:
    Info 0000: Locking the Windows PE image for changes.
    Error 6005: The account is not authorized to log in from this station.
    Info 0000: Stopping service HpcDeployment on node VESTA1
    Info 0000: Updating the configuration of the DHCP scope for the Application network
    Info 0000: Updating the configuration of the DHCP scope for the Private network
    The operation can be identified by the GUID: 5884cd65-8cad-4f18-a93c-58f86b524088. Using this GUID a log of the operation can be obtained from the HPC PowerShell command: Get-HpcOperation -id 5884cd65-8cad-4f18-a93c-58f86b524088 | Get-HpcOperationLog.

    Monday, 23 April 2018 4:15 PM
  • Hi Christoph,

    Could you share the 3 HpcManagement_AA_<index>.bin files with largest indexes on both head node VESTA1 and VESTA2?

    The files locate in C:\Program Files\Microsoft HPC Pack 2016\Data\LogFiles\Management\

    You already know my mail address.

    Thanks,

    Sunbin

    Tuesday, 24 April 2018 8:26 AM