question

If the tables have the same names between the two databases and the databases will be identical after the merge, perhaps you could change the configuration file for your application to only use a single database. This should not necessitate a code change. If you need to continue to use both database, it will depend largely on how your application use the databases. If the tables you will be removing from one database are read-only from the application, you should have a pretty good chance using [dblink][1] (sort of the PostgreSQL version of opendatasource) and replacing the missing tables with views of the same name. If your application writes to the tables you will be removing, the complexity goes up a bit. In addition to views and dblink, you will need to implement [rules on insert, update, and delete][2] to take care of data modifications. I haven't actually implemented either of these scenarios. You can do some further reading on the University of Washington's Network Management Tools [blog][3] about their experience with it. I would also recommend a visit to some [PostgreSQL forums][4] for additional implementation help. [1]: http://www.postgresql.org/docs/8.3/static/contrib-dblink.html [2]: http://www.postgresql.org/docs/8.3/interactive/rules-update.html [3]: http://tools.cac.washington.edu/2010/01/postgresql-fun-cross-database-joins.html [4]: http://www.bing.com/search?q=postgresql+forums