none
Windows 2012 R2 rolling upgrade to 2016 file server RRS feed

  • Question

  • Hi Folks,

        I am not even sure is this fully supported by Microsoft .  I am doing  POC for this one. Sharing with the rest of you of what I have out found also to spare you some time of troubleshooting.

    1. Adding 2016 node must be done in Windows 2016 Failover clustering manager.  Failing to do so cause the cluster to go offline.

    2.In mixed mode ironically, configuring the file server role in Windows 2016 Failover clustering manager will not work .  I need to do it in Windows 2012 node for it to work.

    3. At this point everything seems ok. Both Cluster  and Client Access Point are up.

    4. Problem arises when I want that 2016 node to take ownership of the File server role (need to do this as I want to evict 2012 node one by one). The File server role immediately gone down when I do  this.

    Troubleshooting steps taken:

    1. Delete the Virtual name computer object and create the new one.

    Any help and tips will be appreciated. Thank you.


    • Edited by herngyih Monday, August 19, 2019 11:58 PM
    Monday, August 19, 2019 6:55 AM

All replies

  • You didn't really state what you were attempting to do, so it is a little hard to chime in.

    First, if you are trying to perform a rolling upgrade, you must go from 2012 R2 to 2016.  Can't go from 2012 to 2016.

    Don't understand why you are configuring the file server role on the 2016 node.  That should already be configured on the 2012 R2 node.  You are talking of a rolling upgrade.  That implies the role is already configured.


    tim

    Monday, August 19, 2019 1:09 PM
  • Thank you and sorry if it confused you. Yeah, I was indeed doing a rolling upgrade from 2012 R2 to Win 2016.

    "Don't understand why you are configuring the file server role on the 2016 node.  That should already be configured on the 2012 R2 node.  "

    Yeah you are right. The file server role was configured in the 2012 R2 nodes. But the moment I add a 2016 node and move the ownership to that 2016 node, the file server role goes offline. Therefore, I deleted the file server role and recreate with the 2016 node inside the cluster(mixed mode)  hoping that would works. It didn't.



    • Edited by herngyih Tuesday, August 20, 2019 2:59 AM
    Tuesday, August 20, 2019 12:04 AM
  • HI,

    Thanks for your question.

    Would you like to rolling upgrade the clustered file server from 2012 R2 to 2016?

    Windows Server 2016 provides a new feature——Cluster OS Rolling Upgrade, and it enables an administrator to upgrade the operating system of the cluster nodes

    from Windows Server 2012 R2 to Windows Server 2016 without stopping the Hyper-V or the Scale-Out File Server workloads.

    Here are process of Cluster OS Rolling Upgrade:

    •  Start
    •  Backup Cluster Database
    •  Backup Workload Data
    •  Check CAU is running
    • Pause Node | Drain Roles
    •  Evict Node From the cluster
    • Clean OS install
    •  Join to domain
    • Configure Hyper-V, Storage, Failover Clustering and Virtual Network
    • Add Node to the Cluster and place the workload
    •  Continue these steps with other nodes
    •  Update Cluster Function Level

    l  

    We can refer to the following link for detailed steps,

    Cluster operating system rolling upgrade

    https://docs.microsoft.com/en-us/windows-server/failover-clustering/cluster-operating-system-rolling-upgrade

    https://mssqltrek.com/2017/02/04/upgrading-windows-failover-cluster-2012r2-to-2016/

    Please Note: Since the web site is not hosted by Microsoft, the link may change without notice. Microsoft does not guarantee the accuracy of this information.

    Hope above information can help you.

    Highly appreciate your effort and time. If you have any question or concern, please feel free to let me know.

    Best regards,

    Michael


    Please remember to mark the replies as an answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com



    Tuesday, August 20, 2019 10:34 AM
    Moderator
  • When the file server role went offline, did you first try to bring it back online before removing it from the cluster?

    At this point, if you have the mixed mode cluster without the file service role, it would most likely be easier to finish upgrading to 2016 and then re-add the file service role to the 'new' 2016 cluster rather than going back to restore the 2012 R2 cluster so you can inplace upgrade it.


    tim

    Tuesday, August 20, 2019 1:21 PM
  • "When the file server role went offline, did you first try to bring it back online before removing it from the cluster?"

    I did.

    "At this point, if you have the mixed mode cluster without the file service role, it would most likely be easier to finish upgrading to 2016 and then re-add the file service role to the 'new' 2016 cluster rather than going back to restore the 2012 R2 cluster so you can inplace upgrade it."

    It is not possible to fail over while in mix mode? Also, no upgrade work is involved. All the nodes are VM and basically I am adding a 2016 node to the cluster hoping that the failover works. Then I will evict one 2012 node. Then add the final 2016 node and evict the last 2012 node.

    Wednesday, August 21, 2019 12:03 AM
  • Hi,

    May I confirm your situation again?

    1)Did your cluster nodes all upgrade to 2016? 

    2)Now has File server role been added to the new working cluster? 

    3)Any error message when clicking "show critical events" under the offline and failed resources of FIle server role?

    4)Whether you can failover the file server role to another node?

    5)Please also check if there's any error in the event viewer regarding Failover-Clustering?

    Best regards,

    Michael 


    Please remember to mark the replies as an answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com

    Wednesday, August 21, 2019 9:15 AM
    Moderator
  • Yes, it should be possible to failover in mixed mode.  Doesn't make much sense otherwise.

    "no upgrade work is involved. "

    Don't know what you mean by this statement, as your following description exactly describes an in-place upgrade.  My point is that since something got broken during your first attempt, rather than trying to recreate the original environment, it might be simpler to destroy the original cluster (or whatever components are left) and build a new one.


    tim

    Wednesday, August 21, 2019 12:54 PM
  • "no upgrade work is involved. " What I mean is I didn't upgrade any  of the  2012 nodes to 2016. I add 2016 nodes to the 2012 cluster.

    "My point is that since something got broken during your first attempt, rather than trying to recreate the original environment, it might be simpler to destroy the original cluster (or whatever components are left) and build a new one."

    The objective of my actions is to sort out a efficient methods to upgrade our production cluster. What I have done so far was on test nodes. Destroying cluster and rebuild with new 2016 nodes is not a viable option.  The ideal outcome is just like any other successful example out there, adding 2016 nodes,evict 2012 nodes and raise the cluster functional level. Minimal or no downtime.

    Thursday, August 22, 2019 12:56 AM
  • Okay, if you simply want to 'upgrade' by adding new nodes, then you should have no issues on doing that.  

    https://www.altaro.com/hyper-v/upgrade-2012-r2-cluster-2016-using-cluster-os-rolling-upgrade/ contains a good step-by-step instructions.  Ensure you follow those instructions and see what happens.  Obviously, the only difference would be that you would not need to drain a node and evict it.


    tim

    Thursday, August 22, 2019 1:04 PM
  • Hi,

    Just checking in to see if the information provided was helpful. Please let us know if you would like further assistance.

    Best Regards,

    Michael



    Please remember to mark the replies as an answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com

    Thursday, August 29, 2019 8:14 AM
    Moderator
  • Nope, bottomline is the role just won't  failover to 2016 node.  Still happening even with a newly built 2016 node. The listener IP for dalsapc9v just won't stick to the 2016 node.

    Error 1

    Cluster resource 'dalsapc9v' of type 'Network Name' in clustered role 'dalsapc9v' failed. The error code was '0x80090016' ('Keyset does not exist').

    Based on the failure policies for the resource and role, the cluster service may try to bring the resource online on this node or move the group to another node of the cluster and then restart it.  Check the resource and group state using Failover Cluster Manager or the Get-ClusterResource Windows PowerShell cmdlet.

    Error 2

    Network Name resource 'dalsapc9v' (with associated network name 'DALSAPC9V') has Kerberos Authentication support enabled. Failed to add required credentials to the LSA - the associated error code is '-2146893802'.



    • Edited by herngyih Thursday, September 12, 2019 6:43 AM
    Thursday, September 12, 2019 6:35 AM