I need help to understand session merge() in hibernate. After hours of search in net and books , I am frustrated to understand this .

I have gathered some information though.

merge is used for detached objects .i.e objects which are not attached to DB and disconnected hibernate session.

Now , if we modify the POJO (modifying fields using setter methods) and then again we want to save this POJO , then we need to create a new hibernate session....now do we call merge here or update here ? here I am stuck .

what merge will do ?

what update will do ?

Could you show me a demo code also to use this feature.

This looks complicated to me....I'm very much upset with this concept.

One correction first: Merge is also used for transient instances (Instances which haven't been saved at all yet).

When you have a new hibernate session with detached instance of object A, it doesn't matter in most cases if you use merge() or update() to save the changes. So in your example, calling merge() or update() would yield the exact same result.

However, if you have two objects with the same identifiers where one of them is detached and the other one is persistent, calling update() on the detached instance will throw an exception where merge will save the changes.

Oh, and one more thing. Merge() saves the _STATE_ of the object, so the member variables are saved to the db and the function returns the persistent instance while update() will save the object itself and make it persistent.

Hope this helps :)