When clustering an application for high availability, the preferred configuration is to store the application data on shared storage accessible to all nodes in the cluster. However, VCS also supports the replicated data cluster (RDC) approach, in which the application data is synchronously replicated between nodes. While such configurations tend to be more complex and contain more "moving parts" than the shared storage configuration, they're still supported.
What type of replication are you using? How are you ensuring data consistency between nodes?