@DataJpaTest not updating property annotated with @Version

Question

I have an issue with one test checking that the entity with property version (annotated with @Version) is incremented after an update.

• If I use the annotation @SpringBootTest the version is incremented.
• If I use @DataJpaTest, the version is not incremented.

Do you know how to configure the @DataJpaTest to make the incrementation work?

Please find below the working test:

@SpringBootTest
public class TestAuditing extends TestCrud<AuditingEntity> {

    ...

    @Test
    @Override
    void update() {
        Long version = entity.getVersion();
        super.update();
        assertEquals(++version, entity.getVersion());
    }
}

And here is the not working test:

@ExtendWith(SpringExtension.class)
@DataJpaTest
public class TestAuditing extends TestCrud<AuditingEntity> {
    
    ...

    @Test
    @Override
    void update() {
        Long version = entity.getVersion();
        super.update();
        assertEquals(++version, entity.getVersion());
    }
}

Here is the code of the TestCrud abstract class

public abstract class TestCrud<E extends AbstractEntity> {

    protected static AbstractEntity entity;

    ...

    @Test
    void update() {
        E updatedEntity = crudAdapter.update((E) entity);
        assertEquals(updatedEntity.getId(), entity.getId());
        entity = updatedEntity;
    }
    ...

Thanks in advance for your answer!

Answer 1

I can't give a definitive answer because you didn't posted the code that actually is supposed to do anything. But the likely cause of the problem is that DataJpaTest includes the @Transactioanl annotation, thus moving the transaction boundaries from the entry and exit of the repository method, which I assume you use somewhere to the test method. Since JPA doesn't flush changes until the end of the transaction you don't get a version increment.

For tests like yours which depends on the exact scope of transactions I recommend using a TransactionTemplate, so that you have exact control over your transactions. You can also use SQL logging to ensure that SQL statements get executed when you think they get executed.