用于数据传输对象 (DTO) 投影的 Hibernate 二级缓存

Question

As I have read in many articles (eg here ) - to enable Hibernate's second level cache for given entity we need to set cache concurrency strategy on entity via @org.hibernate.annotations.Cache annotation.

@Entity
@Cache(usage = CacheConcurrencyStrategy.READ_WRITE)
public class Person {

Besides I also use query-level cache (using query.setCacheable(true) ) on some queries that fetches this entity and it works well.

My question relates to custom queries that uses DTO projection, so for the queries like this:

Query query = session.createQuery("SELECT new PersonDto(person.id, person.name) FROM Person person WHERE person.name = :name");
query.setParameter("name", name);
query.setCacheable(true);
query.uniqueResult();

Do I need to set @Cache annotation also for PersonDto ? I have tried to run the query without the annotation and the DTO was successfully cached.

Could you explain why do we need the annotation for entity objects only and other non-entity objects does not require that?

Thanks.

Answer 1

I'm not 100% on this, but you are manually setting cacheable to true for the query.

The annotation on Person is the equivalent for an entity.

I wouldn't think of it as PersonDTO being cached in this instance. If you were to write another query saying select new PersonDTO(person.id, person.name) from Person person where person.id = 10 , I don't think it will look into your cache to see if a PersonDTO with id == 10 exists; whereas, the Entitys cache would because it understands they are the same thing.

I would think of it as the query itself is being cached (meaning if ran again before TTL then cached results would occur). It's caching the fact that you ran this query with a certain name parameter, not that a PersonDTO with that name exists in the cache. Does that make sense?