HI!

Somehow, always my hibrenate throws:
java.sql.BatchUpdateException: Duplicate entry '105-213' for key 1

When the database doesn't have a duplicate record! it seems to me that it's trying to update although i don't ask it to do it.

I thought, as a temporarily solution, to use on duplicate ignore.

How do i ask Hibernate to use it?

As well, what could be a reason to hibernate to insert when i do find?

This exception actually comes from your database,
so you must enable 'ignore duplicate' directly on index on the database (if your database offers this feature at all).



Most probably you are calling a query without changing the default flush mode,
which means that implicitly flush() is called (performing all scheduled insert, update and delete actions).



It's obvious that on the database you never will see duplicate records, because the unique-constraint don't permits it.
But you are trying to insert a record which has the same unique-constraint key as another already inserted record.

what do i need to change in the code?
i simply use:

Do you use Hibernate with Spring? (Unfortunately I don't know Spring).

Anyway: the implicite flush() on the query only anticipates the duplicate key exception.
Disabling the implicite flush() only defers the duplicate key exception to the commit-time
(except you are not intended to commit the transaction at all, but rollback it).
So you have to investigate how to avoid the duplicate insert.

The problem is that i never insert duplicate keys!

I think that hibernate session is not being closed properly, and it thinks that it has new data to insert to the database.

why would it happen?

But somehow something in your code is doing inserts with duplicate keys.
Please enable sql logging either with P6Spy or log4j/jdbc/JDBCAppender,
and report the sql-statements which are executed.
Im sure the concerning sql-statements can be figured out.