Is the node that you are trying to failover to listed as a possible owner for the cluster resources in cluster admin?