Implementation:Apache Shardingsphere JDBCRepository Persist

Knowledge Sources	Apache_Shardingsphere
Domains	Persistence, Standalone, JDBC
Last Updated	2026-02-10 00:00 GMT

Overview

JDBCRepository is a JDBC-backed standalone persist repository using HikariDataSource for hierarchical key-value storage in a relational database table.

Description

JDBCRepository implements StandalonePersistRepository and stores metadata as key-value pairs with parent-child relationships in a single database table. On init, it creates a HikariDataSource from configured JDBC properties (URL, username, password, provider) and executes the provider-specific CREATE TABLE SQL loaded via JDBCRepositorySQLLoader. The query method retrieves a value by exact key match. getChildrenKeys lists child keys under a parent path, extracting the immediate child segment and sorting in reverse order. isExisted checks for key presence. persist recursively creates parent path entries before inserting the leaf value, or updates if the key already exists. update sets a new value for an existing key. delete removes all entries matching the key prefix pattern. The repository supports H2 (in-memory), MySQL, and other JDBC-compatible backends via pluggable SQL providers.

Usage

Use this repository when running ShardingSphere in standalone mode with persistent storage requirements. Configure with type: JDBC and provide JDBC URL, username, password, and provider properties. It is suitable for production standalone deployments requiring durable metadata storage.

Code Reference

Source Location

• Repository: Apache_Shardingsphere
• File: JDBCRepository.java
• Lines: 1-235

Signature

@Slf4j
public final class JDBCRepository implements StandalonePersistRepository {

    public void init(final Properties props)

    public String query(final String key)

    public List<String> getChildrenKeys(final String key)

    public boolean isExisted(final String key)

    public void persist(final String key, final String value)

    public void update(final String key, final String value)

    public void delete(final String key)

    public void close()

    public String getType()
}

Import

import org.apache.shardingsphere.mode.repository.standalone.jdbc.JDBCRepository;

I/O Contract

Inputs

Name	Type	Required	Description
props	Properties	Yes	JDBC connection properties (jdbc_url, username, password, provider)
key	String	Yes	Hierarchical path key (e.g., "/metadata/my_db/rules/sharding")
value	String	Yes	Data value to persist or update

Outputs

Name	Type	Description
queryResult	String	Value stored at the key, or empty string if not found
childrenKeys	List<String>	Immediate child key names under the parent path
isExisted	boolean	Whether the key exists in the repository

Usage Examples

// Initialize with H2 in-memory database
Properties props = new Properties();
props.setProperty("jdbc_url", "jdbc:h2:mem:config;DB_CLOSE_DELAY=-1");
props.setProperty("provider", "H2");
JDBCRepository repository = new JDBCRepository();
repository.init(props);

// Persist hierarchical data
repository.persist("/metadata/my_db/rules/sharding/tables/t_order", yamlContent);

// Query data
String value = repository.query("/metadata/my_db/rules/sharding/tables/t_order");

// List children
List<String> tables = repository.getChildrenKeys("/metadata/my_db/rules/sharding/tables");

// Clean up
repository.close();

Related Pages

• Principle:Apache_Shardingsphere_Metadata_Repository_Persistence