Seed JDBC persistence add-on enables your application to interface with any relational database through the JDBC API.

Show version
dependencies {


You can configure the add-on in one or more of your *.props files. Declare you list of data source names you will be configuring later:

org.seedstack.jdbc.datasources = datasource1, datasource2, ...

Configure each data source separately. Notice the use of the keyword property to specify any property that will be used by the datasource as specific configuration.

provider = HikariDataSourceProvider
driver = org.hsqldb.jdbcDriver
url = jdbc:hsqldb:mem:testdb1
user = sa
password =
property.specific.jdbc.prop = value
property.prop.for.datasource = value

If your app server declares a JNDI datasource:

jndi-name = java:comp/env/jdbc/my-datasource
context = ...

The context property is optional and can be used to specify a particular context name configured in core support to make the lookup. Otherwise the default context will be used.

JDBC Connection

The following examples show how to get a JDBC connection.

public class MyRepository {

    private Connection connection;

    public void updateStuff(int id, String bar){
            String sql = "INSERT INTO FOO VALUES(?, ?)";
            PreparedStatement statement = connection.prepareStatement(sql);
            statement.setInt(1, id);
            statement.setString(2, bar);
        } catch(SqlException e){
            throw new SomeRuntimeException(e, "message");

Any interaction with this connection will have to be realized inside a transaction. Refer to the transaction support documentation for more detail. Below is an example using the annotation-based transaction demarcation (notice the data source name in @Jdbc annotation).

public class MyService {

    private MyRepository myRepository;

    public void doSomethingRelational() {
        myRepository.updateStuff(1, "bar");

Note that the @Jdbc annotation is optional if you have only one type of transactional resources in your application AND if you only have one datasource. If you happen to be in this situation, we still recommend to explicitly specify the annotation to avoid doing so when the project evolves down the road.

DataSource providers

When using a non JNDI datasource, we recommend the use of pooled datasource through a DataSourceProvider defined in the configuration. Three DataSource providers are currently supported out-of-the-box:

We also provide a test oriented DataSource that gives connection directly from the driver. Use PlainDataSourceProvider or do not specify a provider. In case you want to use another data source, you can create your own DataSourceProvider by implementing the DataSourceProvider interface:

public class SomeDataSourceProvider implements DataSourceProvider {

    public DataSource provideDataSource(String driverClass, String
            url, String user, String password, Properties jdbcProperties) {
        SomeDataSource sds = new SomeDataSource();
        return sds;


You will be able to declare it in your configuration as SomeDataSourceProvider (the simple name of your class). Note that if you want to use one of the three datasource providers described above, you will have to add the corresponding dependency to your project.