none
Error Migrating VM -- Network Bandwidth on the Virtualization Host not sufficient

    Question

  • Hi,

    I'm trying to move one of our data servers from one Hyper-V server to another using SCCM VMM 2012 R2.
    The destination server I want to select is unavailable as 'Network Bandwidth on the virtualization host (<HyperV Server Name>) is not sufficient for all network workloads'

    I am able to move other VMs onto the Destination server, and can move VMs off of the Source server. I can move the data server in question to another Hyper V server, just not to the one that I'm after.

    For reference, both the source and destination networks are comparable -- the Management networks consist of 2x 1Gbps NICs teamed in Windows Server 2012 R2. The VM networks are 4x 1Gbps NICs teamed in Windows Server 2012 R2 then provided to to the Hyper Visor as Virtual Switch (as configured in HyperV). Neither the Source or Destination servers networks are close to capacity at the best of times.

    I've done the usual googling and research, however I can find almost nothing on this particular error. There is no useful error message or code to point me in the right direction for troubleshooting. Comparing the configuration of the one HyperV server that VMM will allow the migration for to the one that it will not allow (NIC driver settings, team settings, HyperV settings), I can see no difference at all. 

    Anyone seen and or resolved this error before? Or at the least, point me in a useful direction for troubleshooting?

    Thanks,

    Nick

    

    Tuesday, June 28, 2016 4:37 AM

Answers

  • Hi Nick,

    assuming you are using a logical switch.

    1. Open your VM properties in VMM. Find a vNIC and check the Classification.

    2. Fabric -> Networking -> Logical Switches -> your switch -> properties -> virtual port. Do you see the classification required by the VM?


    http://OpsMgr.ru/

    • Marked as answer by elnickos Tuesday, June 28, 2016 6:57 AM
    Tuesday, June 28, 2016 5:24 AM

All replies

  • Hi Nick,

    assuming you are using a logical switch.

    1. Open your VM properties in VMM. Find a vNIC and check the Classification.

    2. Fabric -> Networking -> Logical Switches -> your switch -> properties -> virtual port. Do you see the classification required by the VM?


    http://OpsMgr.ru/

    • Marked as answer by elnickos Tuesday, June 28, 2016 6:57 AM
    Tuesday, June 28, 2016 5:24 AM
  • Hello,

    Alexey wasn't right, but he did lead me to look in the right direction. While digging around in the 'Fabric' section of VMM I noticed that at some point both the source and destination server were setup as File Shares in VMM. I wasn't able to right-click and remove the servers from here, so I just removed both servers from VMM and re-added them.

    Either it was this, or completely reconfiguring the virtual networks on the adaptors that resolved the issue, I'm not sure. The problem is resolved now though.

    Thanks,

    Nick.

    Tuesday, June 28, 2016 6:57 AM
  • I ran into this problem recently and fixed this issue by:

    - Starting VM migration 

    - Clicking 'Expected Utilization' in the first screen of the 'Migrate VM Wizard'

    - Clicking 'Restore Defaults' (and OK)

    Even though the network bandwidth was filled with a 0, somehow SCVMM does something with the allocated disk space in correlation to Network Bandwidth. Very strange, but this fixed it for me. 

    Monday, June 18, 2018 1:21 PM