Our company uses Oracle Applications for all of its enterprise data (customers, orders, etc). I'm pretty new to Hibernate so please forgive the possible stupid newbie question.

Oracle Apps uses "interface" tables that are used when data is inserted. That is, clients would only insert data into these interface tables and not the "real" tables. So when adding a customer I would write an sql statement that inserts into a table like ra_customers_interface_all. Oracle then processes the data in this table and moves it to the appropriate table(s). When I need to go back and read data, I need to read it from one (or more) of these non-interface tables.

My question is basically....is Hibernate an appropriate solution for something like this considering I am inserting into one table yet reading from another? Additionally, in many cases, I would need to read from BOTH tables in order to aggregate the data into a single object. From what little I know of Hibernate to this point, it's mainly good at mapping objects/attributes to table columns but wouldn't handle a situation where it needs to read from a different location than where it inserted.

You thoughts are appreciated.

map one object A to the "interface table" --> to read, write and modify it
map a second to "source table" B --> only for read, so declare it mutable = false

Also be carefull not to have one instance of A and the corresponding of B into the same hibernate session, you have to manage it well.

Don't use second level cache