I have a group of tables that persist customers' data. At certain points in my application I need to take a snapshot of that data and insert it into an identical but distinct table structure which represents the customer's history trace.

In other words, I have a bunch of tables such as CUSTOMER, APPLICATION, etc. related together and I need the same set of Java objects be able to persist themselves to a different set of tables ie. CUSTOMER_HISTORY, CUSTOMER_APPLICATION etc.
The fields and relations in the _HISTORY tables are identical to the main set of tables so I really don't feel like creating a brand new set of Java classes just to be able to do that. What I need is a clone of the original data in those _HISTORY tables except with all the foreign keys and primary keys regenerated.

Is there a way for Hibernate to dynamically change where its stuff gets persisted or do I need a brand new set of sibling objects mapped to my _HISTORY tables in order to do what I'm trying to do?

All the tables are in the same schema.

Anyone with a suggestion on how to best handle this?

you might be able to do it with a second SessionFactory with different mapping.

In Hibernate 2.2 it will be easier.

Thanks Gavin,

When is 2.2 scheduled for and what's in it that is going to make my problem easier to resolve?

Thanks for a prompt reply.

As usual, there is no schedule.

But not for at least 6 months.

The feature is support for "named" entities, eg.

session.get("AccountLog", id)

Thanks again for your reply. Hibernate is so unbelievably cool I don't want to build another web app without it :-)

The timeline for 2.2 is not alingned with my project's deadline (due in a few weeks) so I think I'll just do it the hard way and have a replica set of Java objects. Still a far better option than messing with triggers ans stored procedures.

Thanks for your work on this wonderful tool.