I have a 8 node cluster with Hyper-V, which will be 10 nodes when it's finally done.
Recently I've been try to add nodes and though that went fine, after about a week I could not open the Failover Cluster Manager anymore.
After some checking I found out that the latest added node was giving problems.
VM's on the node still run and function properly, but most Powershell commands result in a "The requested object does not exist. (Exception from HRESULT: 0x80010114)".
I can suspend the node with Suspend-ClusterNode, but draining roles was unsuccesful in one case.
In the other there were no VM's on the node so suspending went fine.
What I did find out was that when I tried to ping the node from another, proper functioning node, it took a while before the pinging started. It felt like the interface had to come back online on the problem node.
After that, I could add the cluster to the Failover Cluster Manager. However, Powershell commands still give a 0x80010114 error or a CIM error for when I use Get-NetAdapter.
A reboot resolves the problem, but only for about a week.
I know there is a topic with the same title already, but the wbemtest en rollup update "answer" is totally unclear to me why I should change something with wbemtest, or why to install updates that to me have nothing to do with this problem.
Before I did the ping test from a functioning node I pinged my DC and another node from the problem node just fine.
No waiting at all.
The cluster has three networks. Management (host only), Live Migration and iSCSI (also a VMSwitch for certain VM's).
I have no idea where to look. Evenviewer doesn't give me anything I can work with that I can find...