I try to develop a bean that can read more than one DB, but it would be useful (in the earlier time) if read from one DB. And then read all tables (depend of the login and password of the user, maybe the admin). But I don't sure if it can be a good bean. First it would responds to requests of other beans that needs something data from tables belongs a specific DB. For do this, the bean should connect to the DB, then responds, and later terminate the connection. The request bean should, in run time, knows the structure of the tables and then makes something with this data, later this bean should be "clean".
What do you think about it?
Thanks.

getMetadata() method of ResultSet retrieves the number, types and properties of this ResultSet object's columns.