Java: What is the difference between save() and saveAndFlush()?

save() — saves to persistence context. saveAndFlush() — saves and immediately writes to DB.

save():

• Adds to persistence context.
• SQL executed at commit or flush.
• Batch-friendly.
• Returns managed entity.

1@Service
2@Transactional
3public class UserService {
4    public void createUsers() {
5        for (int i = 0; i < 1000; i++) {
6            repo.save(new User("user" + i));
7        }
8        // All 1000 users saved at commit
9    }
10    // COMMIT → single batch INSERT
11}

saveAndFlush():

• Saves and immediately flushes to DB.
• SQL executed immediately.
• Useful for visibility.
• Returns managed entity.

1@Service
2@Transactional
3public class UserService {
4    public void createUserWithImmediateSave() {
5        User user = repo.saveAndFlush(new User("Alice"));
6        // SQL executed immediately
7        System.out.println("ID: " + user.getId());
8    }
9    // COMMIT → no additional SQL
10}

How it works:

1. save() adds the entity to the persistence context.
2. saveAndFlush() adds the entity and immediately flushes to the database.
3. save() defers SQL execution until commit or explicit flush.
4. saveAndFlush() is useful when you need immediate database visibility.

Key difference:

• save() — lazy (batch).
• saveAndFlush() — immediate (SQL now).

Best practice:

• Use save() for batch operations.
• Use saveAndFlush() when immediate visibility is needed.
• Consider using saveAll() for batch inserts.