Spring Boot Data Access

Priority: P0

Implementation Guidelines

JPA & Hibernate (Spring Data JPA)

• Read-Only: Default to

  @Transactional(readOnly = true)

  on Services to optimize DB resources.
• Projections: Use

  Java Records

  for Read-Only query results. Avoid fetching full

  @Entity

  objects when not necessary.
• Pagination: ALWAYS use

  Pageable

  and

  Slice

  (or

  Page

  ) to prevent loading massive datasets.
• Spring Data: Prefer

  JpaRepository

  and

  Query methods

  . Use

  @Query

  with JPQL for complex logic. Use Flyway or Liquibase for migrations; never use

  spring.jpa.hibernate.ddl-auto=create

  (

  ddl-auto

  ) in production.

Query & Transaction Optimization

• N+1 Problem: Fix

  N+1

  selects using

  JOIN FETCH

  (JPQL) or

  @EntityGraph

  .
• Open-In-View: Set

  spring.jpa.open-in-view=false

  in

  application.yaml

  .
• Bulk Operations: Use

  @Modifying

  with

  @Query

  for updates/deletes to bypass EntityManager overhead.
• Connection Pool: Configure

  HikariCP

  with explicit

  maximum-pool-size

  . Tune Hikari pool-size based on expected concurrent queries.

Anti-Patterns

• No N+1 Selects: Use

  JOIN FETCH

  or

  @EntityGraph

  instead of lazy-loading in loops.
• Entity Inflation: Don't use

  @Data

  (Lombok) on Entities as it breaks Proxy and

  hashCode

  /

  equals

  performance.
• Transactional Leak: Don't put

  @Transactional

  on public

  Repository

  methods if the

  Service

  is already transactional.
• Raw SQL: Avoid native SQL unless JPQL/Criteria API is insufficient.

References

• Implementation Examples