Skip to content

SQL Initialization

Automated database schema management and data population on application startup.

FluDa SQL Init resolves versioned SQL scripts from the classpath or file system, tracks which migrations have been applied in a db_migrations history table, and runs pending scripts in order at startup.

Modules

Module Artifact Description
core fluda-sql-init-core Plain SQL script parser and migration runner. Depends only on javax.sql.DataSource.
config fluda-sql-init-config MicroProfile Config integration — populates SqlInitConfig from fluda.sqlinit.* properties.
cdi fluda-sql-init-cdi CDI bean that observes the @Startup event and applies pending migrations automatically.

How it works

  1. Scripts are resolved from the configured locations (default: classpath:db/migration).
  2. Each file must follow the naming convention V<version>__<description>.sql (for example, V1__create_engineers.sql).
  3. On startup, the migrator compares resolved scripts against the db_migrations history table.
  4. Pending migrations are applied in version order, each in its own transaction.
  5. A migration is marked succeeded or failed in the history table.

Supported databases

The history table DDL is provided for the following databases:

Database Auto-detected from JDBC product name or URL
H2 H2
PostgreSQL PostgreSQL, postgres
MySQL MariaDB, MySQL
MSSQL Microsoft SQL Server, sqlserver
Oracle Oracle

The database type is auto-detected from the connection metadata. It can also be set explicitly through SqlInitConfig or the fluda.sqlinit.db-type configuration property.

Reading order

  • Getting Started — add the right dependency, run your first migration, and review configuration (script locations, separator, and database type).
  • Advanced Topics — custom version strategies and resource resolvers.