java - Hibernate:批量和非批量插入生成相同数量的 SQL 查询

Question

有一个代码示例，由两个具有双向映射的实体组成。

@Entity
@Table(name = "Post")
@NoArgsConstructor
@Data
public class Post {
    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
    private long id;

    @OneToMany(mappedBy = "post", cascade = CascadeType.ALL, orphanRemoval = true)
    private Set<Comment> comments = new HashSet<>();

    public void addComment(Comment comment) {
        comments.add(comment);
    }

    public void removeComment(Comment comment) {
        comments.remove(comment);
    }
}

@Entity
@Table(name = "Comment")
@NoArgsConstructor
@Data
@EqualsAndHashCode(exclude = {"id", "post"})
public class Comment {
    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
    private long id;

    @Column(nullable = false)
    private String review;

    @ManyToOne(fetch = FetchType.LAZY)
    @JoinColumn(name = "post_id", updatable = false, nullable = false)
    private Post post;

    public Comment(Post post, String review) {
        this.post = post;
        this.review = review;
    }
}

配置

@SpringBootApplication(exclude = HibernateJpaAutoConfiguration.class)
@PropertySource("classpath:application.properties")
@ComponentScan("jpaDifferent")
public class RootConfig {

    public static void main(String[] args) throws Exception {
        ApplicationContext applicationContext = new AnnotationConfigApplicationContext(RootConfig.class);
        DaoCollections daoCollections = applicationContext.getBean(DaoCollections.class);
        Post post = new Post();
        for (int i = 0; i < 30; i++) {
            post.addComment(new Comment(post, "Comment No. " + i));
        }
        daoCollections.savePost(post);
}

    @Bean
    @ConfigurationProperties(prefix = "db")
    public DataSource dataSource() {
        return new BasicDataSource();
    }

    @Bean
    public LocalSessionFactoryBean sessionFactory() {
        LocalSessionFactoryBean sfb = new LocalSessionFactoryBean();
        sfb.setDataSource(dataSource());
        sfb.setPackagesToScan("jpaDifferent");
        sfb.setHibernateProperties(properties());
        return sfb;
    }

    private Properties properties() {
        Properties properties = new Properties();
        properties.setProperty("hibernate.dialect", "org.hibernate.dialect.MySQL5Dialect");
        properties.setProperty("hibernate.hbm2ddl.auto", "update");
        properties.setProperty("hibernate.show_sql", "true");
        properties.setProperty("hibernate.jdbc.batch_size", "10");
        properties.setProperty("hibernate.order_inserts", "true");
        properties.setProperty("hibernate.order_updates", "true");
        properties.setProperty("hibernate.jdbc.batch_versioned_data", "true");
        return properties;
    }

    @Bean
    public HibernateTransactionManager transactionManager() {
        HibernateTransactionManager transactionManager = new HibernateTransactionManager();
        transactionManager.setSessionFactory(sessionFactory().getObject());
        return transactionManager;
    }
}

道

@Repository
@Transactional
public class DaoCollections {

    @Autowired
    private SessionFactory sessionFactory;

    private Session currentSession() {
        return sessionFactory.getCurrentSession();
    }

    public void savePost(Post post) {
        currentSession().persist(post);
    }
}

Hibernate 日志包含 30 行“插入”:

Hibernate: insert into Post values ( )
Hibernate: insert into Comment (post_id, review) values (?, ?)
Hibernate: insert into Comment (post_id, review) values (?, ?)
...........
Hibernate: insert into Comment (post_id, review) values (?, ?)

根据包含有关批处理的说明的配置文件中的属性，我预计会生成 3 个“插入”SQL 查询。但实际上有 30 个，每个“评论”对应一个。我不知道我是否误解了有关批处理的内容，事实上，它不应该优化查询数量，或者可能只是配置有问题。

Answer 1

这是因为身份ID。当 Id 为 Identity 类型时，Hibernate 不会进行批处理。使用自定义或序列。