最近用到了MyBatis配置多数据源,原以为简单配置下就行了,实际操作后发现还是要费些事的,这里记录下,以作备忘
不多废话,直接上代码,后面会有简单的实现介绍
多数据源时Spring配置文件
<?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:aop="http://www.springframework.org/schema/aop"
xsi:schemaLocation="http://www.springframework.org/schema/beans
http://www.springframework.org/schema/beans/spring-beans-3.0.xsd
http://www.springframework.org/schema/context
http://www.springframework.org/schema/context/spring-context-3.0.xsd
http://www.springframework.org/schema/aop
http://www.springframework.org/schema/aop/spring-aop-3.0.xsd">
<bean id="propertyConfigurer" class="org.springframework.beans.factory.config.PropertyPlaceholderConfigurer">
<property name="location" value="classpath:jdbc.properties"/>
</bean>
<bean id="sqlServerDataSource" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
<property name="driverClassName" value="${jdbc.sqlserver.driver}"/>
<property name="url" value="${jdbc.sqlserver.url}"/>
<property name="username" value="${jdbc.sqlserver.username}"/>
<property name="password" value="${jdbc.sqlserver.password}"/>
<property name="initialSize" value="${jdbc.initialSize}"/>
<property name="minIdle" value="${jdbc.minIdle}"/>
<property name="maxIdle" value="${jdbc.maxIdle}"/>
<property name="maxActive" value="${jdbc.maxActive}"/>
<property name="maxWait" value="${jdbc.maxWait}"/>
<property name="defaultAutoCommit" value="${jdbc.defaultAutoCommit}"/>
<property name="removeAbandoned" value="${jdbc.removeAbandoned}"/>
<property name="removeAbandonedTimeout" value="${jdbc.removeAbandonedTimeout}"/>
<property name="testWhileIdle" value="${jdbc.testWhileIdle}"/>
<property name="timeBetweenEvictionRunsMillis" value="${jdbc.timeBetweenEvictionRunsMillis}"/>
<property name="numTestsPerEvictionRun" value="${jdbc.numTestsPerEvictionRun}"/>
<property name="minEvictableIdleTimeMillis" value="${jdbc.minEvictableIdleTimeMillis}"/>
</bean>
<bean id="mySqlDataSource" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
<property name="driverClassName" value="${jdbc.mysql.driver}"/>
<property name="url" value="${jdbc.mysql.url}"/>
<property name="username" value="${jdbc.mysql.username}"/>
<property name="password" value="${jdbc.mysql.password}"/>
<property name="initialSize" value="${jdbc.initialSize}"/>
<property name="minIdle" value="${jdbc.minIdle}"/>
<property name="maxIdle" value="${jdbc.maxIdle}"/>
<property name="maxActive" value="${jdbc.maxActive}"/>
<property name="maxWait" value="${jdbc.maxWait}"/>
<property name="defaultAutoCommit" value="${jdbc.defaultAutoCommit}"/>
<property name="removeAbandoned" value="${jdbc.removeAbandoned}"/>
<property name="removeAbandonedTimeout" value="${jdbc.removeAbandonedTimeout}"/>
<property name="testWhileIdle" value="${jdbc.testWhileIdle}"/>
<property name="timeBetweenEvictionRunsMillis" value="${jdbc.timeBetweenEvictionRunsMillis}"/>
<property name="numTestsPerEvictionRun" value="${jdbc.numTestsPerEvictionRun}"/>
<property name="minEvictableIdleTimeMillis" value="${jdbc.minEvictableIdleTimeMillis}"/>
</bean>
<bean id="multipleDataSource" class="com.cnblogs.lzrabbit.MultipleDataSource">
<property name="defaultTargetDataSource" ref="mySqlDataSource"/>
<property name="targetDataSources">
<map>
<entry key="mySqlDataSource" value-ref="mySqlDataSource"/>
<entry key="sqlServerDataSource" value-ref="sqlServerDataSource"/>
</map>
</property>
</bean>
<bean id="sqlSessionFactory" class="org.mybatis.spring.SqlSessionFactoryBean">
<property name="dataSource" ref="multipleDataSource"/>
</bean>
<!-- mybatis.spring自动映射 -->
<bean class="org.mybatis.spring.mapper.MapperScannerConfigurer">
<property name="basePackage" value="com.cnblogs.lzrabbit"/>
</bean>
<!-- 自动扫描,多个包以 逗号分隔 -->
<context:component-scan base-package="com.cnblogs.lzrabbit"/>
<aop:aspectj-autoproxy/>
</beans>
MultipleDataSource实现
package com.cnblogs.lzrabbit;
import org.springframework.jdbc.datasource.lookup.AbstractRoutingDataSource;
/**
* Created by rabbit on 14-5-25.
*/
public class MultipleDataSource extends AbstractRoutingDataSource {
private static final ThreadLocal<String> dataSourceKey = new InheritableThreadLocal<String>();
public static void setDataSourceKey(String dataSource) {
dataSourceKey.set(dataSource);
}
@Override
protected Object determineCurrentLookupKey() {
return dataSourceKey.get();
}
}
MyBatis接口Mapper定义,直接使用注解方式实现
public interface MySqlMapper {
@Select("select * from MyTable")
List<Map<String,Object>> getList();
}
public interface SqlServerMapper {
@Select("select * from MyTable")
List<Map<String,Object>> getList();
}
手动数据源切换调用
package com.cnblogs.lzrabbit;
import org.springframework.context.ApplicationContext;
import org.springframework.context.support.ClassPathXmlApplicationContext;
/**
* Created by rabbit on 14-5-25.
*/
public class Main {
public static void main(String[] args) {
//初始化ApplicationContext
ApplicationContext applicationContext = new ClassPathXmlApplicationContext("applicationContext.xml");
MySqlMapper mySqlMapper = applicationContext.getBean(MySqlMapper.class);
SqlServerMapper sqlServerMapper = applicationContext.getBean(SqlServerMapper.class);
//设置数据源为MySql,使用了AOP测试时请将下面这行注释
MultipleDataSource.setDataSourceKey("mySqlDataSource");
mySqlMapper.getList();
//设置数据源为SqlServer,使用AOP测试时请将下面这行注释
MultipleDataSource.setDataSourceKey("sqlServerDataSource");
sqlServerMapper.getList();
}
}
使用SpringAOP方式实现自动切换
package com.cnblogs.lzrabbit;
import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Aspect;
import org.springframework.stereotype.Component;
@Component
@Aspect
public class MultipleDataSourceAspectAdvice {
@Around("execution(* com.cnblogs.lzrabbit.*.*(..))")
public Object doAround(ProceedingJoinPoint jp) throws Throwable {
if (jp.getTarget() instanceof MySqlMapper) {
MultipleDataSource.setDataSourceKey("mySqlDataSource");
} else if (jp.getTarget() instanceof SqlServerMapper) {
MultipleDataSource.setDataSourceKey("sqlServerDataSource");
}
return jp.proceed();
}
}
调用日志
2014-05-25 20:02:04,319 DEBUG [main] com.cnblogs.lzrabbit.MySqlMapper.getList - ooo Using Connection [jdbc:mysql://127.0.0.1:3306/test?useUnicode=true&characterEncoding=UTF-8&allowMultiQueries=true, UserName=root@192.168.1.32, MySQL Connector Java]
2014-05-25 20:02:04,333 DEBUG [main] com.cnblogs.lzrabbit.MySqlMapper.getList - ==> Preparing: select * from MyTable
2014-05-25 20:02:04,371 DEBUG [main] com.cnblogs.lzrabbit.MySqlMapper.getList - ==> Parameters:
2014-05-25 20:02:04,396 DEBUG [main] com.cnblogs.lzrabbit.MySqlMapper.getList - <== Total: 8
2014-05-25 20:02:04,620 DEBUG [main] com.cnblogs.lzrabbit.SqlServerMapper.getList - ooo Using Connection [jdbc:jtds:sqlserver://127.0.0.1:1433/test, UserName=sa, jTDS Type 4 JDBC Driver for MS SQL Server and Sybase]
2014-05-25 20:02:04,620 DEBUG [main] com.cnblogs.lzrabbit.SqlServerMapper.getList - ==> Preparing: select * from TmallCityMap
2014-05-25 20:02:04,621 DEBUG [main] com.cnblogs.lzrabbit.SqlServerMapper.getList - ==> Parameters:
2014-05-25 20:02:04,681 DEBUG [main] com.cnblogs.lzrabbit.SqlServerMapper.getList - <== Total: 397
这里就上面的实现做个简单解释,在我们配置单数据源时可以看到数据源类型使用了org.apache.commons.dbcp.BasicDataSource,而这个代码实现了javax.sql.DataSource接口
配置sqlSessionFactory时org.mybatis.spring.SqlSessionFactoryBean注入参数dataSource类型就是javax.sql.DataSource
实现多数据源的方法就是我们自定义了一个MultipleDataSource,这个类继承自AbstractRoutingDataSource,而AbstractRoutingDataSource继承自AbstractDataSource ,AbstractDataSource 实现了javax.sql.DataSource接口,所以我们的MultipleDataSource也实现了javax.sql.DataSource接口,可以赋值给sqlSessionFactory的dataSource属性
public abstract class AbstractRoutingDataSource extends AbstractDataSource implements InitializingBean {
}
public abstract class AbstractDataSource implements DataSource {
}
再来说下MultipleDataSource的实现原理,MultipleDataSource实现AbstractRoutingDataSource抽象类,然后实现了determineCurrentLookupKey方法,这个方法用于选择具体使用targetDataSources中的哪一个数据源
<bean id="multipleDataSource" class="com.cnblogs.lzrabbit.MultipleDataSource">
<property name="defaultTargetDataSource" ref="mySqlDataSource"/>
<property name="targetDataSources">
<map>
<entry key="mySqlDataSource" value-ref="mySqlDataSource"/>
<entry key="sqlServerDataSource" value-ref="sqlServerDataSource"/>
</map>
</property>
</bean>
可以看到Spring配置中multipleDataSource设置了两个属性defaultTargetDataSource和targetDataSources,这两个属性定义在AbstractRoutingDataSource,当MyBatis执行查询时会先选择数据源,选择顺序时现根据determineCurrentLookupKey方法返回的值到targetDataSources中去找,若能找到怎返回对应的数据源,若找不到返回默认的数据源defaultTargetDataSource,具体参考AbstractRoutingDataSource的源码
public abstract class AbstractRoutingDataSource extends AbstractDataSource implements InitializingBean {
private Map<Object, Object> targetDataSources;
private Object defaultTargetDataSource;
/**
* Retrieve the current target DataSource. Determines the
* {@link #determineCurrentLookupKey() current lookup key}, performs
* a lookup in the {@link #setTargetDataSources targetDataSources} map,
* falls back to the specified
* {@link #setDefaultTargetDataSource default target DataSource} if necessary.
* @see #determineCurrentLookupKey()
*/
protected DataSource determineTargetDataSource() {
Assert.notNull(this.resolvedDataSources, "DataSource router not initialized");
Object lookupKey = determineCurrentLookupKey();
DataSource dataSource = this.resolvedDataSources.get(lookupKey);
if (dataSource == null && (this.lenientFallback || lookupKey == null)) {
dataSource = this.resolvedDefaultDataSource;
}
if (dataSource == null) {
throw new IllegalStateException("Cannot determine target DataSource for lookup key [" + lookupKey + "]");
}
return dataSource;
}
/**
* Determine the current lookup key. This will typically be
* implemented to check a thread-bound transaction context.
* <p>Allows for arbitrary keys. The returned key needs
* to match the stored lookup key type, as resolved by the
* {@link #resolveSpecifiedLookupKey} method.
*/
protected abstract Object determineCurrentLookupKey();
.............
}
在动态切换数据源方法时选择了AOP方式实现,这里实现的简单粗暴,具体应用时根据实际需要灵活变通吧
题外话,这里提下SqlServer驱动选择的问题,目前SqlServer的驱动主要有微软的官方驱动和JTDS驱动两种,关于这两个驱动我做过测试,批量更新,在小数据量(100以下)时,JTDS相对微软驱动性能稍微高一点点,在数据量增大时几万到上百万时,微软驱动有着明显优势,所以若对性能比较敏感,建议使用微软驱动,否则随意
微软驱动在Maven库找不到,这点比较郁闷,若使用maven的话还得先安装到本地,这点很不爽
JTDS使用比较方便Maven直接引用即可
相关jar maven引用
View Code