On Twitter I asked the following to David James (Director of Engineering, ConfigMgr, Microsoft) and Johan Arwidmark (CTO @ TrueSec):
Multicast during an SCCM 2012 R2 SP1 (1511 release) Task Sequence fails with error “Failed to get MCS key (Code 0x80004005)”. This error is found in the smsts.log log file on the (Windows 10 Enterprise x64 1511) client machine.
With TPM 1.2, Microsoft was able to clear the TPM during the SCCM Task Sequence without asking for permission to clear the TPM. With TPM 2.0, SCCM is unable to clear and activate the TPM chip during the deployment. The first time you boot your computer, you need to provide a BitLocker Recovery Key, or the tpm.msc console will tell you that the TPM is ready for use, with reduced functionality.
Because I wanted to configure Device Guard with Windows 10, I need the Hyper-V Hypervisor to be enabled on Windows 10. I tried to do this with DISM and an answer file, but it’s not possible to enable Hyper-V during the Task Sequence Deployment because Hyper-V requires a couple of reboots.
Create a new “Set Task Sequence Variable” task in your Task Sequence. This will run the PowerShell command after the Task Sequence ends. I’ve set this task before enabling the Driver Package, but it should be possible to place this task anywhere you like.
Task Sequence Variable: SMSTSPostAction
Value: powershell -ExecutionPolicy ByPass -Command “Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V-Hypervisor -all -NoRestart;Disable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V-Tools-All,Microsoft-Hyper-V-Services -NoRestart”
This will do the following:
- Enable all the Hyper-V Features after the deployment
- Remove the Hyper-V Tools and Services (Management Tools) afterwards. I found out that this is the best way to only add the Hyper-V Hypervisor.
You still need to reboot the system a few times to enable this feature. Because I enabled the BitLocker PIN, I can’t reboot the machine because it will ask for a PIN a few times.
Recently I connected System Center – Virtual Machine Manager with WSUS. The WSUS server is installed on the primary site server of my SCCM 2012 R2 SP1 CU2 installation. After I configured my SCCM WSUS server as an update server for VMM, the distribution point in the office stopped working. You will see HTTP ERROR “12030” in your logs and the PXE request on a client will fail. Browsing to the website of the SCCM Primary Site server will fail too.
I found out that the certificate of IIS on my primary site was gone. There was no certificate selected for the Default Website. After adding the certificate again and restarting IIS, PXE started to work again.
Recently I found the following error in the SMSPXE.log log file on my newly created distribution point:
CryptVerifySignature failed, 80090006 SMSPXE <REMOVED TIME> 2500 (0x09C4)
untrusted certificate: <REMOVED CERTIFICATE> SMSPXE <REMOVED TIME> 2500 (0x09C4)
Failed to get information for MP: https://SCCMPRIMARY.DOMAIN.TLD. 80090006. SMSPXE <REMOVED TIME> 2500 (0x09C4)
After recreating my certificate template for the IIS Service on the primary site server, it fixed the problem. Check the online documentation of SCCM for the details of this certificate template.
Sometimes it’s possible that the registry keys SerializedMCSKey and SignedSerializedMCSKey in the HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\SMS\MCS location are empty after a fresh installation or after reinstalling multicast.
Patience… It took like 5 or 6 hours to get those values populated by SCCM / WDS. I’ve searched for a way to force this, but I couldn’t find anything online. I’ve tried to reboot both machines, without any luck. If you know a way to force this, please let me know.