Is it possible to do a Windows 2008 R2 hyper-v Enterprise in place upgrade to Windows 2012 R2 hyper-v Datacenter? RRS feed

  • Question

  • Is it possible to do a Windows 2008 R2 hyper-v Enterprise in place upgrade to Windows 2012 R2 hyper-v Datacenter?

    We have lots of VMs on the windows 2008 R2 hyper-v server and we are looking for a way to upgrade safely.

    Can we do an in place upgrade to windows 2012 R2 with the VM's shut down on the hyper-v? 

    Or do we need to move these VMs to another location during this upgrade?

    We also have FC SAN storage on the hyper-v server and so am wondering if the emulex cards(PCI fiber cards) would require reinstalling the firmware?

    Also would netapp DSM/MPIO software have to be reinstalled after this upgrade?

    Please provide a good reference link to support your recommendations.


    • Edited by kimdav111 Saturday, February 25, 2017 4:02 AM adding another question
    Saturday, February 25, 2017 3:55 AM

All replies

  • Technically you could do an in-place upgrade but not recommended. If something goes wrong during the upgrade you could possibly end up with nothing left. Better to clean install and migrate.



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

    Saturday, February 25, 2017 4:07 AM
  • I agree with Dave. A clean install is always recommended, but an upgrade is possible.

    An upgrade should not impact the storage, so other than the fact that it is recommended practice to backup before any change, there is no need to worry about the VMs on the storage.

    As to requirements for the Emulex cards and NetApp DSM, those questions have to be answered by the respective vendors. Microsoft does not control those aspects - the owners of the hardware/software do.

    . : | : . : | : . tim

    Saturday, February 25, 2017 2:27 PM
  • Been looking a articles which make statement of not recommending an in place upgrade.  What are some of the specific missteps that can occur during an in place upgrade?  


    Saturday, February 25, 2017 6:31 PM
  • Agree with Tim. Backups are not for cowards.

    Depending on DSM and drivers, I would recommend you to reinstall them after upgrade in any way without trying to use them before. They could be compatible with new OS, but who knows which .dll were changed in driver packages for 2016 version?

    Found this for your last question:


    It's pretty general, but in 1st paragraph the're definitely right. No one knows which combination of server roles you've applied to your system.

    Great phrase: "one way that always works is a clean installation

    My personal advice is to perform a clean installation if your server works only as Hyper-V server (you didn't install Domain Controller role, I suppose?). Thus you'll achive a most clean OS installation. Those drivers DSM modules and virtual switches can be recreated and reinstalled with a little effort. BUT. You'll have a much less chance that new OS installation (compared to upgrade) will impact all your VMs. 

    Saturday, February 25, 2017 7:22 PM
  • "specific missteps " are usually user error, so it is next to impossible to create a list of things that users do.  Fat finger can cause a lot of pain.  I remember issuing a del *.* when logged into a system account.

    The engineers have tried to make upgrades simple, and for the most part they are painless.  The recommendation for clean install is the key word 'clean'.  Your new system has no artifacts from any previous configurations that had been done on the original system.  When it comes to debugging an issue in the future, you don't have to deal with any issues that may have lingering effects as a result of an in-place upgrade.

    The time it takes to perform a clean installation on something like Hyper-V and the time it takes to perform an in-place upgrade is about the same.  Yes, there would be a little extra time required to import the VMs, but that can easily be scripted so you don't have to sit there watching it until it completes.

    . : | : . : | : . tim

    Sunday, February 26, 2017 4:58 PM