The Pain of JDBC and the Persistence Context
Compare JDBC and JPA, and verify the first-level cache, identity, change detection, and write-behind through logs.
- Reproduce the first-level cache, identity, change detection, and write-behind respectively within the same transaction.
- Record the expected SQL, the actual SQL, and the timing of their occurrence for each experiment.
- Compare this with the state management that must be handled manually when processing the same requirements using JDBC.
Submission Evidence · Verification tests by function · SQL logs showing execution order · Explanation compared to JDBC · Answers to evidence-based questions 1–4
함께 답할 근거형 질문 4개
- Why is the first-level cache not a global application cache?
- Within what scope is the identity of entities with the same identifier guaranteed, and how did you verify it?
- At what point does dirty checking compare which information to generate an UPDATE statement?
- How did you prove with logs the point where write-behind changes the actual SQL execution order?


