I don't want to start a long discussion about inplace upgrades of Servers, I never did it in the past 15 years as well, but now with the Server 2012 R2 I'm confident and would like to do our first inplace upgrade of our 2008 R2 with File Server role to 2012 R2.
The existing 2008R2 has 3 Thick Provisioned vDisks. vDisk 1 has 60 GB and has Windows 2008 R2 on it. vDisk 2 and vDisk 3 are both 1.9TB and spanned into 1 Dynamic NTFS Volume with 3.8TB in Windows where all the shares and files are located.
In my mind now is:
1. Create a VM Backup and a snapshot only of the vDisk 1 to be prepared for a worst case
2. Do the inplace upgrade
3. Check if everything went smoth, all settings and so on.
But if something should go wrong, and this is my question, could I just restore the Snapshot of the vDisk1 with OS 2008R2, Reboot the server and it will recognize the spanned Data Volume with the 2 other vDisks again and everything is back to normal? Or would it work to delete the vDisk1 physically on the datastore and restore the vDisk1 from the Backup and reboot again the old 2008R2 OS? My concerns are just if the upgrade changed something on the patition tables of the Data Volumes, will the OS recognize the spanned Data Volume afterwards again?
I hope it was understandable what I mean 😉
Thx for any recommendations.