读写分离

概念

为了缓解数据库压力,将写入和读取操作分离为不同数据源,写库称为主库,读库称为从库,一主库可配置多从库。

支持项

  1. 提供了一主多从的读写分离配置,可独立使用,也可配合分库分表使用。
  2. 独立使用读写分离支持SQL透传。
  3. 同一线程且同一数据库连接内,如有写入操作,以后的读操作均从主库读取,用于保证数据一致性。
  4. Spring命名空间。
  5. 基于Hint的强制主库路由。

不支持范围

  1. 主库和从库的数据同步。
  2. 主库和从库的数据同步延迟导致的数据不一致。
  3. 主库双写或多写。

代码开发示例

仅读写分离

// 构建读写分离数据源, 读写分离数据源实现了DataSource接口, 可直接当做数据源处理. masterDataSource, slaveDataSource0, slaveDataSource1等为使用DBCP等连接池配置的真实数据源
Map<String, DataSource> dataSourceMap = new HashMap<>();
dataSourceMap.put("masterDataSource", masterDataSource);
dataSourceMap.put("slaveDataSource0", slaveDataSource0);
dataSourceMap.put("slaveDataSource1", slaveDataSource1);

// 构建读写分离配置
MasterSlaveRuleConfiguration masterSlaveRuleConfig = new MasterSlaveRuleConfiguration();
masterSlaveRuleConfig.setName("ms_ds");
masterSlaveRuleConfig.setMasterDataSourceName("masterDataSource");
masterSlaveRuleConfig.getSlaveDataSourceNames().add("slaveDataSource0");
masterSlaveRuleConfig.getSlaveDataSourceNames().add("slaveDataSource1");

DataSource dataSource = MasterSlaveDataSourceFactory.createDataSource(dataSourceMap, masterSlaveRuleConfig);

分库分表 + 读写分离

// 构建读写分离数据源, 读写分离数据源实现了DataSource接口, 可直接当做数据源处理. masterDataSource0, slaveDataSource00, slaveDataSource01等为使用DBCP等连接池配置的真实数据源
Map<String, DataSource> dataSourceMap = new HashMap<>();
dataSourceMap.put("masterDataSource0", masterDataSource0);
dataSourceMap.put("slaveDataSource00", slaveDataSource00);
dataSourceMap.put("slaveDataSource01", slaveDataSource01);

dataSourceMap.put("masterDataSource1", masterDataSource1);
dataSourceMap.put("slaveDataSource10", slaveDataSource10);
dataSourceMap.put("slaveDataSource11", slaveDataSource11);

// 构建读写分离配置
MasterSlaveRuleConfiguration masterSlaveRuleConfig0 = new MasterSlaveRuleConfiguration();
masterSlaveRuleConfig0.setName("ds_0");
masterSlaveRuleConfig0.setMasterDataSourceName("masterDataSource0");
masterSlaveRuleConfig0.getSlaveDataSourceNames().add("slaveDataSource00");
masterSlaveRuleConfig0.getSlaveDataSourceNames().add("slaveDataSource01");

MasterSlaveRuleConfiguration masterSlaveRuleConfig1 = new MasterSlaveRuleConfiguration();
masterSlaveRuleConfig1.setName("ds_1");
masterSlaveRuleConfig1.setMasterDataSourceName("masterDataSource1");
masterSlaveRuleConfig1.getSlaveDataSourceNames().add("slaveDataSource10");
masterSlaveRuleConfig1.getSlaveDataSourceNames().add("slaveDataSource11");

// 通过ShardingSlaveDataSourceFactory继续创建ShardingDataSource
ShardingRuleConfiguration shardingRuleConfig = new ShardingRuleConfiguration();
shardingRuleConfig.getMasterSlaveRuleConfigs().add(masterSlaveRuleConfig0);
shardingRuleConfig.getMasterSlaveRuleConfigs().add(masterSlaveRuleConfig1);

DataSource dataSource = ShardingDataSourceFactory.createDataSource(dataSourceMap, shardingRuleConfig);

Spring命名空间配置示例

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xmlns:context="http://www.springframework.org/schema/context"
    xmlns:sharding="http://shardingjdbc.io/schema/shardingjdbc/sharding"
    xmlns:masterslave="http://shardingjdbc.io/schema/shardingjdbc/masterslave"
    xsi:schemaLocation="http://www.springframework.org/schema/beans 
                        http://www.springframework.org/schema/beans/spring-beans.xsd
                        http://www.springframework.org/schema/context 
                        http://www.springframework.org/schema/context/spring-context.xsd 
                        http://shardingjdbc.io/schema/shardingjdbc/sharding 
                        http://shardingjdbc.io/schema/shardingjdbc/sharding/sharding.xsd 
                        http://shardingjdbc.io/schema/shardingjdbc/masterslave 
                        http://shardingjdbc.io/schema/shardingjdbc/masterslave/master-slave.xsd 
                        ">
    <!-- 配置真实数据源 -->
    <bean id="dbtbl_0_master" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
        <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
        <property name="url" value="jdbc:mysql://localhost:3306/dbtbl_0_master"/>
        <property name="username" value="root"/>
        <property name="password" value=""/>
    </bean>
    
    <bean id="dbtbl_0_slave_0" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
        <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
        <property name="url" value="jdbc:mysql://localhost:3306/dbtbl_0_slave_0"/>
        <property name="username" value="root"/>
        <property name="password" value=""/>
    </bean>
    
    <bean id="dbtbl_0_slave_1" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
        <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
        <property name="url" value="jdbc:mysql://localhost:3306/dbtbl_0_slave_1"/>
        <property name="username" value="root"/>
        <property name="password" value=""/>
    </bean>
    
    <bean id="dbtbl_1_master" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
        <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
        <property name="url" value="jdbc:mysql://localhost:3306/dbtbl_1_master"/>
        <property name="username" value="root"/>
        <property name="password" value=""/>
    </bean>
    
    <bean id="dbtbl_1_slave_0" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
        <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
        <property name="url" value="jdbc:mysql://localhost:3306/dbtbl_1_slave_0"/>
        <property name="username" value="root"/>
        <property name="password" value=""/>
    </bean>
    
    <bean id="dbtbl_1_slave_1" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
        <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
        <property name="url" value="jdbc:mysql://localhost:3306/dbtbl_1_slave_1"/>
        <property name="username" value="root"/>
        <property name="password" value=""/>
    </bean>
    
    <!-- 定义读写分离数据源, 读写分离数据源实现了DataSource接口, 可直接当做数据源处理 -->
    <master-slave:data-source id="dbtbl_0" master-data-source-name="dbtbl_0_master" slave-data-source-names="dbtbl_0_slave_0, dbtbl_0_slave_1" strategy-type="ROUND_ROBIN" />
    <master-slave:data-source id="dbtbl_1" master-data-source-name="dbtbl_1_master" slave-data-source-names="dbtbl_1_slave_0, dbtbl_1_slave_1" strategy-type="ROUND_ROBIN" />
    
    <sharding:inline-strategy id="databaseStrategy" sharding-column="user_id" algorithm-expression="dbtbl_${user_id % 2}" />
    <sharding:inline-strategy id="orderTableStrategy" sharding-column="order_id" algorithm-expression="t_order_${order_id % 4}" />
    
    <sharding:data-source id="shardingDataSource">
        <sharding:sharding-rule data-source-names="dbtbl_0, dbtbl_1">
            <sharding:table-rules>
                <sharding:table-rule logic-table="t_order" actual-data-nodes="dbtbl_${0..1}.t_order_${0..3}" database-strategy-ref="databaseStrategy" table-strategy-ref="orderTableStrategy"/>
            </sharding:table-rules>
        </sharding:sharding-rule>
    </sharding:data-source>
</beans>

使用Hint强制路由主库示例

HintManager hintManager = HintManager.getInstance();
hintManager.setMasterRouteOnly();
// 继续JDBC操作