Implementation:Apache Shardingsphere StatisticsStorageEngine Store

Knowledge Sources	Apache_Shardingsphere
Domains	Cluster, Statistics
Last Updated	2026-02-10 00:00 GMT

Overview

StatisticsStorageEngine computes row-level statistics differences between current and changed data and persists the adds, updates, and deletes.

Description

StatisticsStorageEngine is a final class constructed with a ContextManager, database name, schema name, table name, and a collection of new row/column values. The storage method retrieves the current table schema, creates RowStatistics from the new values using RowStatisticsCollectorUtils, and compares them against the current statistics obtained from the ShardingSphereStatistics hierarchy. It produces an AlteredDatabaseStatistics containing three lists: added rows (keys present in changed but not current), updated rows (keys present in both but with different values), and deleted rows (keys present in current but not changed). The diff is serialized via YamlRowStatisticsSwapper and persisted through StatisticsPersistService.update.

Usage

Use this engine when new statistics data has been collected for a specific table and needs to be diff-merged into the persistent store. It is called by StatisticsRefreshEngine after collecting row/column values from a DialectDatabaseStatisticsCollector.

Code Reference

Source Location

• Repository: Apache_Shardingsphere
• File: StatisticsStorageEngine.java
• Lines: 1-103

Signature

@RequiredArgsConstructor
public final class StatisticsStorageEngine {

    public StatisticsStorageEngine(final ContextManager contextManager,
                                   final String databaseName, final String schemaName,
                                   final String tableName,
                                   final Collection<Map<String, Object>> rowColumnValues)

    public void storage()
}

Import

import org.apache.shardingsphere.mode.manager.cluster.statistics.StatisticsStorageEngine;

I/O Contract

Inputs

Name	Type	Required	Description
contextManager	ContextManager	Yes	Provides access to database metadata, statistics, and persist facade
databaseName	String	Yes	Target database name
schemaName	String	Yes	Target schema name
tableName	String	Yes	Target table name
rowColumnValues	Collection<Map<String, Object>>	Yes	Newly collected row data as column name-value maps

Outputs

Name	Type	Description
void	void	Side effects: persists added, updated, and deleted row statistics to the repository

Usage Examples

// After collecting row/column values from a statistics collector:
Collection<Map<String, Object>> rowColumnValues = collector.collectRowColumnValues(
    databaseName, schemaName, tableName, metaData).orElse(Collections.emptyList());

// Create the storage engine and persist the diff
new StatisticsStorageEngine(contextManager, databaseName, schemaName, tableName, rowColumnValues)
    .storage();

Related Pages

• Principle:Apache_Shardingsphere_Cluster_Metadata_Synchronization