Spring Boot Data Access

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