2023-08-03T12:16:13.789Z

2023-08-04T08:18:21.093Z

0

Hibernate: 
    select
        post0_.id as id1_2_0_,
        user1_.user_id as user_id1_4_1_,
        post0_.create_at as create_a2_2_0_,
        post0_.modified_at as modified3_2_0_,
        post0_.content as content4_2_0_,
        post0_.is_deleted as is_delet5_2_0_,
        post0_.title as title6_2_0_,
        post0_.user_id as user_id7_2_0_,
        user1_.email as email2_4_1_,
        user1_.name as name3_4_1_,
        user1_.oauth_id as oauth_id4_4_1_,
        user1_.picture as picture5_4_1_,
        user1_.role as role6_4_1_ 
    from
        post post0_ 
    inner join
        users user1_ 
            on post0_.user_id=user1_.user_id 
    order by
        post0_.id desc limit ?,
        ?

저도 그렇게 생각하는데 Pageable 을 사용하지 않고 .fetct() 로 List<Post> 를 반환하니 count 쿼리가 따로 나가지 않아서요.. 그래서 찾아보니 fetch() 로 조회하면 count 쿼리가 따로 나가질 않으니 필요하다면 애플리케이션 로직에서 처리하던가, count 쿼리를 따로 호출해야한다고 하더라고요. 근데 count 쿼리가 나가질 않는데 어떻게 페이지네이션이 동작하는지 그걸 모르겠습니다..

0

Pageable로 호출한 부분 코드를 보여주실 수 있을까여?

0

// Controller
@GetMapping("/posts")
  public ResponseEntity<PostMultiReadResponse> readAll(@ModelAttribute PageSearch pageSearch) {
    PostMultiReadResponse response = postService.readAll(pageSearch);

    return ResponseEntity.ok(response);
  }

// Service
@Transactional
  public PostMultiReadResponse readAll(PageSearch pageSearch) {
    List<Post> offset = postRepository.readAll(pageSearch);
//    List<Post> noOffset = postRepository.readNoOffset(pageSearch);

    List<PostSingleReadResponse> response = offset.stream()
        .map(post -> PostSingleReadResponse.builder()
            .id(post.getId())
            .title(post.getTitle())
            .content(post.getContent())
            .createdDate(post.getCreateAt())
            .modifiedDate(post.getModifiedAt())
            .writer(WriterResponse.builder()
                .name(post.getUser().getName())
                .email(post.getUser().getEmail())
                .picture(post.getUser().getPicture())
                .build())
            .build())
        .collect(Collectors.toList());

    return PostMultiReadResponse
        .builder()
        .postsResponse(response)
        .build();
  }

// Repository
@Override
  public List<Post> readAll(PageSearch pageSearch) {
    return jpaQueryFactory.selectFrom(QPost.post)
        .join(QPost.post.user)
        .fetchJoin()
        .limit(pageSearch.getSize())
        .offset(pageSearch.getOffset())
        .orderBy(QPost.post.id.desc())
        .fetch();
  }

// Service Test
  @Test
  @DisplayName("게시글 전체 조회 API 페이징 성능을 테스트합니다.")
  void readAll() {
    // given
    PageSearch pageSearch = new PageSearch(1, 20);

    // when
    long startTime = System.currentTimeMillis();
    List<Post> response = postRepository.readAll(pageSearch);
    long endTime = System.currentTimeMillis();
    long pagingTime = endTime - startTime;
    System.out.println("반환 시간: " + pagingTime + "ms");

    // then
    assertThat(response).isNotNull();
  }

넵 차례대로 Controller, Service, Repository, Test 코드입니다.

테스트 코드 실행 결과는 아래와 같습니다.

Hibernate: 
    select
        post0_.id as id1_2_0_,
        user1_.user_id as user_id1_4_1_,
        post0_.create_at as create_a2_2_0_,
        post0_.modified_at as modified3_2_0_,
        post0_.content as content4_2_0_,
        post0_.is_deleted as is_delet5_2_0_,
        post0_.title as title6_2_0_,
        post0_.user_id as user_id7_2_0_,
        user1_.email as email2_4_1_,
        user1_.name as name3_4_1_,
        user1_.oauth_id as oauth_id4_4_1_,
        user1_.picture as picture5_4_1_,
        user1_.role as role6_4_1_ 
    from
        post post0_ 
    inner join
        users user1_ 
            on post0_.user_id=user1_.user_id 
    order by
        post0_.id desc limit ?,
        ?

반환 시간: 177ms