System proxy databases behave like user proxy databases, with some notable enhancements and exceptions. System proxy databases are only used in an HA configuration.
System proxy databases allow customer-written applications to run on either node in a high-availability cluster. This does not imply “single-system image” capability; rather, it suggests an environment in which most user-written applications can execute on either node in the cluster. This means that both databases and user-created objects should be visible to both nodes.
A system proxy database has the same name as the database in the primary node it references, and contains handling for the user-defined objects that are necessary to support the application. Proxy tables are created for each user table and view found in the primary database, and stored procedures are converted to RPCs and forwarded to the node referenced by the proxy database.
