I am following the procedure GCP defined here: https://cloud.google.com/compute/docs/instances/windows/creating-windows-persistent-disk-snapshot#read-only
THE ISSUE:
After I create a new disk from my snapshot and CLEAR its READONLY flag (per these instructions above), I am stuck with the following ERROR in the GCP SERIAL CONSOLE and the new VM refused to boot.
It looks like the new VM “hardware” changed enough to make windows hate me. I have selected all the identical virtual machine settings.
Has anyone been successful at this?
My overall intent is to provide a way to clone a production Windows Server into our test project or roll back when a patch/upgrade “boogers” up the server.
I am not sure why it took me so long to find it, but I found this article: https://learn.microsoft.com/en-us/troubleshoot/azure/virtual-machines/windows/error-code-0xc000000e
It turns out that the Boot Configuration Data (BCD) had been messed up during the shadowcopy process. Why? Dunno…
But using the command :
bcdedit /store D:\boot\bcd /enum /v
I am able to see several of the “device” value were set to unknown.
then using the following command where I replace with “Windows Boot Loader”. Identifier AND with “D” (drive letter of drive after it was mounted for repair). After the changes, I was able to attach this drive to a NEW GCP VM and it booted.
bcdedit /store :\boot\bcd /set {bootmgr} device partition=:
bcdedit /store :\boot\bcd /set {bootmgr} integrityservices enable
bcdedit /store :\boot\bcd /set {} device partition=:
bcdedit /store :\boot\bcd /set {} integrityservices enable
bcdedit /store :\boot\bcd /set {} recoveryenabled Off
bcdedit /store :\boot\bcd /set {} osdevice partition=:
bcdedit /store :\boot\bcd /set {} bootstatuspolicy IgnoreAllFailures
