I recently had an issue where one of my Hyper-V nodes (Let's call it Node D) in a failover cluster would fail it's health check and consistently reboot every 30-40 min. Logs stated that the problems were with the netft driver, which after led me to DCOM errors with the vmms service. I tried to re-register the service, but that only caused the service to disappear completely. I was left with having to uninstall and re-install the Hyper-V role on the node. That went without apparent issue, but now whenever I boot Node D (and do not disable all but the management NIC), all the other nodes (Nodes A-E) in my cluster lose connectivity to the CSV shared storage and all my virtual machines cannot be reached! When I shut down the Node D, all connectivity returns.
I've disabled the NICs (except management), and even removed the iSCSI connections to the storage on Node D. It can't talk to the storage. What is it on Node D that is causing my other nodes to lose connectivity? I think it might be Failover Clustering, but what is Node D doing to be taking the storage away from the other nodes?