I’ve run into this with increasing frequency in my Server 2012R2 clusters, and it ended up being WMI corruption. IPConfig would show correct IP addressing (and even the teamed NICs created with Windows teaming), and ping would work; however, the network connections dialog box (ncpa.cpl) was completely empty, Server Manager said teaming was disabled, launching the teaming administration page (lbfoadmin.exe) would show “Host Unmanageable” as the status, and attempting to make any connection to any other machine other than ping would result in failure (for example, \\server\share would fail immediately with the host being unavailable).
Here are the steps I needed to use to get the machines up and running again (and I’ve restored from failed backups on a few to test and retest these steps, and in this particular order which did seem to matter, to validate):
1. If this is a Hyper-V server, remove all virtual switches (make note of settings, and recreate-reattach once the server has been fixed and teams rebuilt)
– You may need to mark the vSwitch as private, and reboot, before you are able to successfully remove the vSwitch
– You may need to uncheck the Microsoft NDIS Capture extension before being able to remove the vSwitch
2. Install ndiscleanup (You will NEED to get this from Microsoft … or search the interwebs…) and run from an elevated cmd prompt:
ndiscleanup_x64.exe -cleanup -all
(this can take many minutes, but will likely clean up an interface or two once it finishes)
3. Completely rebuild WMI – this is usually verboten, but in the case of complete repository corruption, this is required to restore WMI functionality:
– Take ownership of, and then rename, the folder “Repository” under C:\Windows\System32\wbem\ to “Repository.old” – you will need to stop and disable the winmgmt (Windows Management Instrumentation) service before you can do this.
– Run the following script:
@echo off
winmgmt /standalonehost
sc config winmgmt start= disabled
net stop winmgmt /y
%systemdrive%
cd %windir%\system32\wbem
for /f %%s in ('dir /b *.dll') do regsvr32 /s %%s
wmiprvse /regserver
winmgmt /regserver
sc config winmgmt start= Auto
net start winmgmt
for /f %%s in ('dir /s /b *.mof *.mfl') do mofcomp %%s
4. Delete the following from the registry:
Key: HKLM\System\CurrentControlSet\Control\Network Value: Control
Key: HKLM\System\CurrentControlSet\Services\NdisImPlatform\Parameters
– Delete *all subkeys*, do not delete “Parameters” key itself
5. Open an administrative-level cmd prompt, and run the following commands:
Set devmgr_show_nonpresent_devices=1 Devmgmt.msc
– Click View > Show Hidden Devices, and delete any “ghosted” unavailable network interface cards listed
– Uninstall *all* physical network adapters
6. Reset the IP stack from an elevated cmd prompt:
netsh int ip reset netsh int ipv4 reset netsh int ipv6 reset
7. Reboot, and log on as the local admin
8. Open an elevated PowerShell prompt, and try to execute the following commands – if they run without error (even if they return no data), the network stack should be working properly at this point (if you receive any WMI class error from these commands, start over from #2 – I have seen instances where this has been done 2 – 3 times before everything is “fixed” – still trying to figure that one out….):
Get-NetAdapter Get-NetAdapterBinding Get-NetLbfoTeam
9. At this point, ncpa.cpl and lbfoadmin should both come back “healthy”, showing network adapters properly (even though any network teams previously created will be gone) as well as lbfo status as “Online” (rather than “Host Unmanageable”). Running ipconfig from a cmd prompt should also return back “healthy” (although adapters may need static IP addresses re-configured).