We had a SAN failure this weekend on our 2008R2 Hyper-V Failover Cluster, this dropped all VM’s within the cluster and left us with a few VM’s unable to boot. In failover cluster manager, selecting the VM displayed the status as failed, and selecting the cluster node that hosted the VM shows its status as pending as the below screenshots show:
To resolve this, the Hyper-V node which was currently the owner of the failed VM’s had to be rebooted, this required live migrating several virtual servers before it could be done. Once rebooted, the VM’s were available to be booted on any node within the cluster. It appears that the node was preventing the VM from booting, perhaps it maintained a lock on the config/VHD files?