To use C-JDBC, databases are replicated to a set of servers, and C-JDBC is configured to know each element in this database cluster. Reads passed through C-JDBC are then load-balanced across each node, with writes being broadcast to each node to maintain synchronicity.
Among other changes, the 1.2 release supports Updatable ResultSets. New pass-through schedulers have been added for all RAIDb levels for increased performance. The DatabaseMetaData code has been refactored to provide more transparency with regard to the underlying database engine. A number of other contributions have helped in various area for developer documentation, code cleaning and bug fixes.
This project enables a middleware component to replicate many features of a database cluster, which might be very useful if a formal DBA isn't available or the database product being used doesn't support clustering.
What do you think?"