Do You Really Need Hibernate?
summary
Argues that full JPA/Hibernate is unnecessary for data‑heavy CRUD apps and demonstrates replacing it with jOOQ plus Java Records, testcontainers and Flyway. Shows generating metadata from a real DB, handling an ERP with ~2,000 tables and ~5,000 procedures that produced 11,000 generated classes, and writing one SQL using multiset/JSON to return a five‑level ranking tree. Covers upserts, returning autogenerated columns, batching, debug logging of prepared+inlined SQL, performance tradeoffs and how jOOQ can be combined with JPA when needed.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:29Introducing Juke
- 3:58Entity state vs data set
- 5:25Why no JPA/Hibernate
- 6:27SQL vs ORM scalability
- 8:13Juke origins and licensing
- 9:58Nested data necessity
- 12:46Generating Juke code
- 15:42Juke plugin config
- 17:07Metadata classes and active record
- 21:23Testing Juke with Spring Boot
- 22:48Inserting and updating records
- 25:15Projections and joins in Juke
- 27:22Type-safe projections with DTOs
- 28:51Deleting records and returning fields
- 30:35Loading nested data
- 34:31Spring projections and N+1
- 36:01Ranking lists with nested queries
- 40:23Demo and generated SQL
- 44:23Juke vs JPA/Hibernate
- 45:45Q&A
related talks
43mDo You Really Need Hibernate?
54mDo you really need Hibernate
54mDo you really need Hibernate?
46mThe State of Java Relational Persistence
51mThe Persistence Heavyweight Championship: JPA vs. jOOQ
50mJOOQ Data Abstractions Without Distraction
50mCreate Your Perfect Persistence Stack
47mThere is always a choice. Spring Data JDBC vs. Hibernate. Taras Boychuk. Java Fest 2020 Online