ddl-auto options in yaml

Overview

ddl-auto is a setting that tells your framework (like Spring Boot) how to manages the schema (the tables and columns)

• Imagine a robot carpenter lol
• When your application starts, the robot looks at your blueprints (your code) and then looks at the house (your PostgreSQL database).

Options

• create
  • Creates schema on startup (destroys existing data!)
  • drops all existing tables and creates brand new ones from scratch based on your code
  • Great for tests, terrible for production
• create-drop
  • Creates schema on startup, drops on shutdown (testing only)
  • Same as create, but drops all the tables when the application shuts down. Only used for testing
• update
  • Checks the database against your code → If a table is missing, it creates it. If a column is missing, it adds it. It will not delete columns, as that could cause data loss.
  • most common one for development
• validate
  • Only validates that schema matches entities (safe)
  • looks at the database tables and compares them to your code. If they don’t match perfectly, the application will fail to start
  • Used to make sure that your database schema is correct
• none
  • Does absolutely nothing with schema
  • the robot carpenter is turned off lol

In production

• validate
  • best option
  1. Fail fast - Find problems at startup, not in production traffic
  2. Schema drift protection - Ensures your code matches your database
  3. Zero risk - Never modifies your database, only checks it
• none
  • unless u have specific reasons to use this use validate
  • why u might want to use this:
    • Database is managed separately (DBA team handles all schema changes)
    • Using database migrations (like Flyway/Liquibase) and want zero JPA interference
    • Legacy system where JPA entities don’t perfectly match database