Private Cloud Simulator FAQ

.

Please send email to pvsha@microsoft.com for any queries related to PCS or WSSD, AzureStack hardware certification.

If you encounter failures and need help, open a bug (feedback) in the Collaborate portal, attach the required HLKX files, and inform the PCS help team.

If the test failures are caused by Microsoft, the team will resolve the bug as fixed. You can later use the bug number to get an errata for your submission.

• Go to Microsoft Collaborate site to open a new feedback bug
• Select EEAP | General
• "Product" field is set to "Windows Server", "Product Family" field is set to “Windows Server”, and "Partner Feature Division" field is set to “SDDC - Certification, Azure Stack”.
• Attach your files to the bug.

Before running PCS test, make sure initial pagefile size is greater than 50GB. A dump file wont get created if pagefile size is too small. See Advanced troubleshooting for Stop error or blue screen error issue.

By default, Windows automatically manages paging file size and its location. Default location is C:\pagefile.sys and file size could grow while test runs.

If your boot disk of cluster node has very little free space due to a large pagefile, you can change pagefile location and size by modifying the PagingFiles value. Below is an expample that set the initial and maximum pagefile sizes to 50GB (51200MB).

reg add "HKLM\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management" /v PagingFiles /t REG_MULTI_SZ /d "C:\pagefile.sys 51200 51200" /f

Because compute cluster nodes normally have a lot of memory to host VMs, by default, PCS use active dump setting to reduce the size of memory dump file. More info can be found here. If you need to change your environments to use ‘complete memory dump’ setting, follow the steps below to collect complete memory dump files. Make sure you have enough free disk spaces and your page file size is bigger than the total memory size.

• Go to HLK controller machine
• Open file "C:\Program Files (x86)\Windows Kits\10\Hardware Lab Kit\Tests\amd64\Pcs\PrivateCloudSimulator-Manager.psm1"
• Find the regkey FilterPages in the file and change the value from 1 to 0
  • From: & "reg.exe" add "HKLM\SYSTEM\CurrentControlSet\Control\CrashControl" /v FilterPages /t REG_DWORD /d 1 /f
  • To: & "reg.exe" add "HKLM\SYSTEM\CurrentControlSet\Control\CrashControl" /v FilterPages /t REG_DWORD /d 0 /f
• Rerun the test.

HLK update packages can be downloaded from Microsoft Collaborate site.

• To certify Windows Server 2016: https://partner.microsoft.com/dashboard/collaborate/packages/3959
• To certify Windows Server 2019: https://partner.microsoft.com/dashboard/collaborate/packages/5485

When the HLK update package ZIP file is downloaded from internet, the file may contains a ‘block’ attribute. You MUST unblock the ZIP file before extracting the files. You can use PowerShell Unblock-File cmdlet to unblock a file. Below is an example,

PS> Unblock-File filename

To update HLK controller, do the following steps on the HLK controller.

• Make sure HLK Studio and HLK Manager are closed.
• Copy the unblocked HlkUpdatePackage zip file to HLK controller.
• Unzip the ZIP file.
• Open an elevated PowerShell window and run '.\UpdateHlkController.ps1' script.
  • PS > .\UpdateHlkController.ps1

The HLK update pacakge for Windows Server 2016 contains a script to update a HLK client. If you are trying to certify SAS-attached devices for Windows Server 2016, you should run this script on each HLK client. This script updates the storage gatherer on HLK clients to not select the firmware update feature for SAS-attached devices by default. Should a customer or partner seek the firmware update AQ, the feature must be manually added during device selection.

• Copy the unblocked HlkUpdatePackage zip file to each HLK client.
  • You have to copy the unblocked files to each HLK client machine. The update script cannot be run remotely.
• Unzip the ZIP file.
• Open an elevated PowerShell window and run '.\UpdateHlkClient.ps1' script.

  PS > .\UpdateHlkClient.ps1

• Restart the client machine to finish update

• If you are trying to certify a network adapter device for Windows Server 2019, you have to disable .NET strong name verification on the machine under the test. Otherwise, Tests related to NDISTest may fail to load test assemblies.
• For testing purpose, you can use the commands below to completely disable strong name verification.
  • reg DELETE "HKLM\Software\Microsoft\StrongName\Verification" /f
  • reg ADD "HKLM\Software\Microsoft\StrongName\Verification*,*" /f

.

Error message: System.TimeoutException: Timed out while trying to reboot <MachineName>

• Go to MachineName
• Verify that secure boot is disabled.
  • Confirm-SecureBootUEFI cmdlet returns false.
• Verify that TESTSIGNING is enabled.
  • Run BCDEDIT.exe tool. TESTSIGNING is listed as YES under current Windows Boot Loader.
• Verify that Bugchecker.exe works.
  • A folder called c:\LmTestBinaries should already exist on the node. If not copy bugchecker files from PCS machine. Exampel: xcopy \MachineName\c$\pcs\bugchecker* c:\LmTestBinaries
  • After running the following command, machine should bluescreen and auto restart itself. If machine doesn't bluescreen, your files may be corrupted.
  • c:\LmTestBinaries\Bugchecker.exe bugcheck

.

Error message: System.Net.WebException: There was an error downloading 'http://localhost/ReportServer/ReportService2010.asmx'.

• Verify MSSQLServer service is running
• If not running, manually start it. If it stops automatically, check if you see an error event in the Application event log saying the "SQL server evaluation period has expired".
• If the evaluation period has expired, you should rebuild a new PCS controller and rerun the test. It will install a new evaluation SQL server. Otherwise, you need to upgrade the SQL Server with a valid SQL Server license key.

.

.

• Open LaunchCreateExportVM_DataTime.log file
• If you see the following error message, please make sure you have all the required KB files and listed them in the correct order.
  • Creating Export VM failed
  • Fail to exit KB Package install process in 360 minutes
  • System.Management.Automation.RuntimeException: Fail to exit KB Package install process in 360 minutes
  • Removing Vm
• Most of the Windows Update Packages require you to install 'servicing stack update (SSU)' first. For example,
  • KB4501371
  • In "How to get this update", it says KB4504369 is required.
    • if you are using Windows Update, the latest SSU (KB4504369) will be offered to you automatically. To get the standalone package for the latest SSU, search for it in the Microsoft Update Catalog.
  • In other words, when running the job, you need to eneter two KBs in test parameter KBPackagePath.
  • example: 'c:\KB\Windows-KB4504369-x64.msu,c:\KB\Windows-KB4501371-x64.msu'
• If KBs are listed in the correct, go to PCS machine and verify that you can install the KBs in the VM manually.

• Verify that you have correct hash values for your VHD files.
  • For Windows Server 2016, see the "Links to the required files" table in Private Cloud Simulator for Windows Server 2016
  • For Windows Server 2019, see the "Links to the required files" table in Private Cloud Simulator for Windows Server 2019
• Verify that you can consistently hit the issue by running the job directly.
  • Go to HLK machine
  • Open "HLK Manager"
  • Click "Explorers" => "Job Explorer"
  • On the left panel, select feature "$\Logo Jobs\System\Solutions\PrivateCloudSimulator"
  • One the right panel, double click "Create Export VM for SDN" job
  • Click "Schedule" button
    • In Jobs tab, enter VHDSoucePath and KBPackagePath
      • Example for Windows Server 2016:
        • VHDSourcePath: c:\pcs\BaseVHDX\14393.0.amd64fre.rs1_release.160715-1616_server_serverdatacentereval_en-us.vhdx
        • KBPackagePath: 'c:\KB\Windows-KB4503537-x64.msu,c:\KB\Windows-KB4503294-x64.msu'
    • In "Machines" tab:
      • Select the machine pool that contains PCS machine
      • Check "Schedule Jobs on the Machines selected from this Machine Pool"
      • Check the checkbox for PCS machine
  • Click "Create Schedule" button
  • After this step, test is scheduled to run. Use the steps below to monitor job status
  • Click "Explorers" => "Job Monitor"
  • On left panel, select the machine pool that contains PCS machine
  • You should see job status and results in the right panel.