none
Multiple processor group device test appears to not wait for reboot RRS feed

  • Question

  • I am attempting to certify a KMDF driver with HCKStudio8_1 on a Server 2012 client.

    The Multiple processor group device test fails on the "validate multiple groups exist step" with the error

    Cause : Failed to Start the Task

    Cause : Failed to Get the user Token

    Failure : Failed to Start the Task "Validate multiple groups exist"

    Cause : Failed to Get User Token ""\"" For Task "Validate multiple groups exist"

    Cause : No Active LoggedOn Session Found.

    The groupsize.wtl file actually says the test passed and reports no errors.

    The WttEa.log file contains the line 

    "2400 1792 2014:1:2 20:53:14:318 Error: 0x4dd, The operation being requested was not performed because the user has not logged on to the network. The specified service does not exist.   (Error ( status = 1245 ) in call: dwRet = GetUserSession (NULL, NULL, &dwSessionID, pUserToken, TRUE) )   File=d:\wtt25\wtt25\sdktools\wtt\jobs\runtime\wttexecutionagent\common\session\src\session.cpp Line=412"

    It appears to have been attempting to do this while the client was still rebooting.

    Any thoughts?

    Thursday, January 2, 2014 10:25 PM

All replies

  • hi Kurt,

    are trying to do creating processor group manually? in HCK 2.1, test only create and delete the processor group, so no manual intervention is required.

    Thanks,

    Mudit

    Friday, January 3, 2014 11:04 AM
  • I'm letting the test create the group.  When the client comes up the groupsize has been changed but the controller has already decided the test has failed and doesn't run the steps to put it back.
    Friday, January 3, 2014 5:00 PM
  • Try to delete all group size manually before running the test. The test might be failing because of existing group size. if possible, can you post the error message here.

    Thanks,

    Mudit

    Monday, January 6, 2014 5:59 AM
  • I have done that.  In fact, I have to do that to put the client back in that state since the test fails and stops running before it gets to the step to restore the groupsize.  Is there a way to make the test wait longer for the client to reboot?
    Monday, January 6, 2014 4:35 PM
  • I can get the test to pass if I pause the virtual machine the test studio is running on while the client reboots following changing the groupsize.  The issue appears to be the client is taking longer to configure itself following the first reboot than the test expects.
    Monday, January 6, 2014 10:47 PM
  • Oh that the twist. I was about to ask this,  is Virtual Machine is installed on SUT. Multiple processor group usually don't pass if Hyper-V role is installed on the SUT.
    Tuesday, January 7, 2014 4:39 AM
  • HyperV is not installed on the client.  The controller+studio is on a virtual machine running under ESXi.

    Tuesday, January 7, 2014 6:44 PM
  • Kurt, HCK is not really supported on Virtual machines. See link here on pre-requisites - http://msdn.microsoft.com/en-us/library/windows/hardware/jj124068.aspx

    >>>>I can get the test to pass if I pause the virtual machine the test studio is running on while the client reboots >>>>following changing the groupsize

    Though I have to see the logs to be sure, based on your comments it does sound like that's the case.

     

    Tuesday, January 7, 2014 7:14 PM