A lot of stuff relating to clustering is specific to a particular container. For WebLogic, when a server-side client obtains a
JDBC Connection via a JNDI lookup and a DataSource, a local object is returned by the DataSource from the Connection pool it references (rather than going to another cluster member for its JDBC connections). The Connection is pinned to the local server instance for the duration of the database transaction, and as long as the application code retains it (until the Connection is closed).
So, clustering the Connections does not enable failover of Connections but it can ease the process of reconnection when a Connection fails.