Spring Data JPA

Entities

@Entity
@Table(name = "transactions")
public class Transaction {

    @Id
    @GeneratedValue(strategy = GenerationType.IDENTITY)
    private Long id;

    private String description;

    @Column(nullable = false, precision = 12, scale = 2)
    private BigDecimal amount;

    private LocalDate date;

    @ManyToOne(fetch = FetchType.LAZY)
    @JoinColumn(name = "category_id")
    private Category category;

    @ManyToOne(fetch = FetchType.LAZY)
    @JoinColumn(name = "user_id", nullable = false)
    private User user;

    @CreationTimestamp
    private Instant createdAt;
}

Repositories — Almost No Code Needed

public interface TransactionRepository extends JpaRepository<Transaction, Long> {

    List<Transaction> findByUserIdAndCategoryId(Long userId, Long categoryId);

    List<Transaction> findByDateBetween(LocalDate start, LocalDate end);

    @Query("SELECT SUM(t.amount) FROM Transaction t WHERE t.user.id = :userId")
    BigDecimal sumAmountByUserId(@Param("userId") Long userId);
}

JpaRepository<Entity, IdType> gives you save, findById, findAll, deleteById, pagination, and more — for free.

Query Derivation Rules (method-name queries)

Spring Data parses method names into queries: findBy + field + optional keyword (And, Or, Between, LessThan, OrderBy...Desc, Containing, IgnoreCase).

List<Transaction> findByAmountGreaterThanAndDateBefore(BigDecimal amount, LocalDate date);

Use this for simple queries; switch to @Query (JPQL) or @Query(nativeQuery = true) once the method name gets unreadable or you need complex joins/aggregations.

Relationships

@OneToMany / @ManyToOne — default fetch for @ManyToOne is EAGER; always override to FetchType.LAZY unless you have a specific reason, to avoid unnecessary joins and N+1 problems.
@ManyToMany — usually better modeled as two @OneToMany/@ManyToOne relationships through an explicit join entity, giving you room to add metadata (e.g. created_at on the join row) later.

The N+1 Query Problem

Lazy-loading a collection inside a loop triggers one query per iteration. Fix with:

JOIN FETCH in a JPQL query, or
@EntityGraph on the repository method, or
batch fetching (spring.jpa.properties.hibernate.default_batch_fetch_size).

@Query("SELECT t FROM Transaction t JOIN FETCH t.category WHERE t.user.id = :userId")
List<Transaction> findAllWithCategory(@Param("userId") Long userId);

`@Transactional`

@Service
public class TransactionService {

    @Transactional
    public void transferBetweenAccounts(Long fromId, Long toId, BigDecimal amount) {
        // all-or-nothing: an exception rolls back every change made in this method
    }
}

Put @Transactional on service methods, not repositories or controllers. By default it rolls back on unchecked exceptions only — use rollbackFor = Exception.class if you need checked exceptions to roll back too.

Spring Data JPA

Spring Data JPA

Entities

Repositories — Almost No Code Needed

Query Derivation Rules (method-name queries)

Relationships

The N+1 Query Problem

@Transactional

On this page

`@Transactional`