Error applying Network Adapter changes on Live Migration or Hyper-V Replica

Recently I setup a new Windows Server 2016 Server with Hyper-V in my home lab environment.  I configured Hyper-V Replica to replicate some of the VMs from the older Windows Server 2012R2 system over to the new 2016 Server.  Replication went fine and today I attempted to failover to the new system so I could repurpose the old server hardware.  Planned Failover went fine and when I went to connect the Network cards of the VMs on the new server, I was greeted with the error below “Error applying Network Adapter changes”.

This error persisted with any of the VMs I had replicated to the new server.  I attempted to do a live migration of one of the non replicated VMs and got a similar error.

I jumped into the event viewer on the Server 2016 system and found the following two errors occurred when I was attempting to modify the Network Card settings.

Log Name:      System
Source:        Microsoft-Windows-Hyper-V-VmSwitch
Date:          1/07/2017 12:36:10 PM
Event ID:      82
Task Category: None
Level:         Error
Keywords:
User:          N/A
Computer:      DL380G7-2.domain.local
Description:
Failed to complete bandwidth policy operation ‘Policy Set’ on port FD853DBF-62EC-4B6C-B696-8DF50C5EB0B9 (Friendly Name: Dynamic Ethernet Switch Port) on switch 76E7111F-32C0-491E-88F5-8ACE9274CB32 (Friendly Name: Home LS) Reservation – 0, Weight – 10, Limit – 0, BurstLimit – 0, BurstSize – 0, Reason – The switch’s bandwidth reservation mode does not support this bandwidth configuration, Status = The request is not supported..

Log Name:      System
Source:        Microsoft-Windows-Hyper-V-VmSwitch
Date:          1/07/2017 12:36:10 PM
Event ID:      78
Task Category: None
Level:         Error
Keywords:
User:          SYSTEM
Computer:      DL380G7-2.domain.local
Description:
Failed to complete operation ‘Policy Add’ on port FD853DBF-62EC-4B6C-B696-8DF50C5EB0B9 (Friendly Name: Dynamic Ethernet Switch Port) on switch 76E7111F-32C0-491E-88F5-8ACE9274CB32 (Friendly Name: Home LS) Property Id {24ad3ce1-69bd-4978-b2ac-daad389d699c} Instance Id {07148245-ac3f-446a-b34d-f48aa0b67138}. Status = The request is not supported..

I did some further digging using the get-vmswitch powershell command and found that the underlying Virtual Switch had differences on the BandwidthReservationMode setting.  The old server had this BandwidthReservationMode setting configured as Weight, whereas the new server had this setting BandwidthReservationMode  configured as Absolute.  Now I’m not 100% certain why they are different, it could be related to the fact that I configured the old servers Hyper-V switches using SCVMM 2012R2 and the new 2016 servers Hyper-V switches were configured locally using the Hyper-V MMC.  Either way, I didn’t want to reconfigure the Hyper-V switch as there were machines running already on the new server.

I did some further investigation on the VMs on the new server and compared their network settings using the Get-VMNetworkAdapter “VMNAME” |FL.  I then found that the key difference was the MinimumBandwidthWeight setting.  Below you can see the failing machine.

I then figured to try to change that setting using the Set-VMNetworkAdapter “VMNAME” –MinimumBandwidthWeight 0 command from Powershell.

After running that command on the machine the setting changed as you can see below.

I was then able to reconnect the VM to the local Hyper-V switch and got things up and running.  I need to do some further investigation into this as it challenges my thoughts on the ease with which we should be able to move VMs around using either Live Migration or Hyper-V replica, but for now I’ve solved this issue.

Hope this saves someone else the challenges I had with it today.