JDBC architecture: drivers, connections & the four core interfaces

How JDBC abstracts database access and the interfaces every Java developer needs to know.

What JDBC is

JDBC (Java Database Connectivity) is a standard API in java.sql that lets Java code talk to any relational database — MySQL, PostgreSQL, SQLite, Oracle — through a common interface. The database vendor ships a JDBC driver (a jar) that implements that interface. Your application code never calls the driver directly; it talks to the standard API.

Connecting to a database

You open a connection via DriverManager.getConnection, passing a JDBC URL, a username, and a password:

Connection conn = DriverManager.getConnection(
    "jdbc:postgresql://localhost:5432/mydb",
    "alice",
    "s3cret"
);

JDBC URL format: jdbc:<subprotocol>://<host>:<port>/<database>[?options]

Subprotocol	Example URL
`postgresql`	`jdbc:postgresql://localhost:5432/mydb`
`mysql`	`jdbc:mysql://localhost:3306/mydb`
`h2`	`jdbc:h2:mem:testdb` (in-memory)
`sqlite`	`jdbc:sqlite:/path/to/file.db`

The four core interfaces

Interface	Role
`Driver`	Registers with `DriverManager`; loaded from the jar
`Connection`	Represents one database session; source of statements
`Statement` / `PreparedStatement`	Executes SQL
`ResultSet`	Holds rows returned by a query

You rarely touch Driver directly — the driver jar auto-registers itself via a ServiceLoader mechanism when it's on the classpath (since JDBC 4.0 / Java 6). Just add the driver jar, write the URL, and DriverManager finds the right driver.

Your code only ever names the java.sql interfaces. Changing the JDBC URL and swapping the driver jar changes the bottom two bands and nothing above them.