Spring Boot 集成Mybatis实现主从（多数据源）分离方案

新建一个Maven项目，最终项目结构如下：

多数据源注入到sqlSessionFactory
POM增加如下依赖：

<!--JSON-->
        <dependency>
            <groupId>com.fasterxml.jackson.core</groupId>
            <artifactId>jackson-core</artifactId>
        </dependency>
        <dependency>
            <groupId>com.fasterxml.jackson.core</groupId>
            <artifactId>jackson-databind</artifactId>
        </dependency>
        <dependency>
            <groupId>com.fasterxml.jackson.datatype</groupId>
            <artifactId>jackson-datatype-joda</artifactId>
        </dependency>
        <dependency>
            <groupId>com.fasterxml.jackson.module</groupId>
            <artifactId>jackson-module-parameter-names</artifactId>
        </dependency>

        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-jdbc</artifactId>
        </dependency>
        <dependency>
            <groupId>mysql</groupId>
            <artifactId>mysql-connector-java</artifactId>
        </dependency>
        <dependency>
            <groupId>com.alibaba</groupId>
            <artifactId>druid</artifactId>
            <version>1.0.11</version>
        </dependency>
        <!--mybatis-->
        <dependency>
            <groupId>org.mybatis.spring.boot</groupId>
            <artifactId>mybatis-spring-boot-starter</artifactId>
            <version>1.1.1</version>
        </dependency>
        <!--mapper-->
        <dependency>
            <groupId>tk.mybatis</groupId>
            <artifactId>mapper-spring-boot-starter</artifactId>
            <version>1.1.0</version>
        </dependency>
        <!--pagehelper-->
        <dependency>
            <groupId>com.github.pagehelper</groupId>
            <artifactId>pagehelper-spring-boot-starter</artifactId>
            <version>1.1.0</version>
            <exclusions>
                <exclusion>
                    <artifactId>mybatis-spring-boot-starter</artifactId>
                    <groupId>org.mybatis.spring.boot</groupId>
                </exclusion>
            </exclusions>
        </dependency>

这里需要注意的是：项目是通过扩展mybatis-spring-boot-starter的org.mybatis.spring.boot.autoconfigure.MybatisAutoConfiguration来实现多数据源注入的。在mybatis-spring-boot-starter：1.2.0中，该类取消了默认构造函数，因此本项目依旧使用1.1.0版本。需要关注后续版本是否会重新把扩展开放处理。
之所以依旧使用旧方案，是我个人认为开放扩展是合理的，相信在未来的版本中会回归。

增加主从库配置(application.yml)

druid:
    type: com.alibaba.druid.pool.DruidDataSource
    master:
        url: jdbc:mysql://192.168.249.128:3307/db-test?characterEncoding=UTF-8&autoReconnect=true&zeroDateTimeBehavior=convertToNull&useUnicode=true
        driver-class-name: com.mysql.jdbc.Driver
        username: root
        password: root
        initial-size: 5
        min-idle: 1
        max-active: 100
        test-on-borrow: true
    slave:
        url: jdbc:mysql://192.168.249.128:3317/db-test?characterEncoding=UTF-8&autoReconnect=true&zeroDateTimeBehavior=convertToNull&useUnicode=true&characterEncoding=utf-8
        driver-class-name: com.mysql.jdbc.Driver
        username: root
        password: root
        initial-size: 5
        min-idle: 1
        max-active: 100
        test-on-borrow: true

创建数据源

@Configuration
@EnableTransactionManagement
public class DataSourceConfiguration {

    @Value("${druid.type}")
    private Class<? extends DataSource> dataSourceType;

    @Bean(name = "masterDataSource")
    @Primary
    @ConfigurationProperties(prefix = "druid.master")
    public DataSource masterDataSource(){
        return DataSourceBuilder.create().type(dataSourceType).build();
    }

    @Bean(name = "slaveDataSource")
    @ConfigurationProperties(prefix = "druid.slave")
    public DataSource slaveDataSource1(){
        return DataSourceBuilder.create().type(dataSourceType).build();
    }
}

将多数据源注入到sqlSessionFactory中
前面提到了这里通过扩展mybatis-spring-boot-starter的org.mybatis.spring.boot.autoconfigure.MybatisAutoConfiguration来实现多数据源注入的

@Configuration
@AutoConfigureAfter({DataSourceConfiguration.class})
public class MybatisConfiguration extends MybatisAutoConfiguration {

    private static Log logger = LogFactory.getLog(MybatisConfiguration.class);

    @Resource(name = "masterDataSource")
    private DataSource masterDataSource;
    @Resource(name = "slaveDataSource")
    private DataSource slaveDataSource;

    @Bean
    public SqlSessionFactory sqlSessionFactory() throws Exception {
        return super.sqlSessionFactory(roundRobinDataSouceProxy());
    }

    public AbstractRoutingDataSource roundRobinDataSouceProxy(){
        ReadWriteSplitRoutingDataSource proxy = new ReadWriteSplitRoutingDataSource();
        Map<Object,Object> targetDataResources = new ClassLoaderRepository.SoftHashMap();
        targetDataResources.put(DbContextHolder.DbType.MASTER,masterDataSource);
        targetDataResources.put(DbContextHolder.DbType.SLAVE,slaveDataSource);
        proxy.setDefaultTargetDataSource(masterDataSource);//默认源
        proxy.setTargetDataSources(targetDataResources);
        return proxy;
    }
}

实现读写分离（多数据源分离）
这里主要思路如下：
1-将不同的数据源标识记录在ThreadLocal中
2-通过注解标识出当前的service方法使用哪个库
3-通过Spring AOP实现拦截注解并注入不同的标识到threadlocal中
4-获取源的时候通过threadlocal中不同的标识给出不同的sqlSession

标识存放ThreadLocal的实现

public class DbContextHolder {

    public enum DbType{
        MASTER,SLAVE
    }

    private static final ThreadLocal<DbType> contextHolder = new ThreadLocal<>();

    public static void setDbType(DbType dbType){
        if(dbType==null)throw new NullPointerException();
        contextHolder.set(dbType);
    }

    public static DbType getDbType(){
        return contextHolder.get()==null?DbType.MASTER:contextHolder.get();
    }

    public static void clearDbType(){
        contextHolder.remove();
    }

}

注解实现

/**
 * 该注解注释在service方法上，标注为链接slaves库
 * Created by Jason on 2017/3/6.
 */
@Target({ElementType.METHOD,ElementType.TYPE})
@Retention(RetentionPolicy.RUNTIME)
public @interface ReadOnlyConnection {
}

Spring AOP对注解的拦截

@Aspect
@Component
public class ReadOnlyConnectionInterceptor implements Ordered {

    public static final Logger logger = LoggerFactory.getLogger(ReadOnlyConnectionInterceptor.class);

    @Around("@annotation(readOnlyConnection)")
    public Object proceed(ProceedingJoinPoint proceedingJoinPoint,ReadOnlyConnection readOnlyConnection) throws Throwable {
        try {
            logger.info("set database connection to read only");
            DbContextHolder.setDbType(DbContextHolder.DbType.SLAVE);
            Object result = proceedingJoinPoint.proceed();
            return result;
        }finally {
            DbContextHolder.clearDbType();
            logger.info("restore database connection");
        }
    }

    @Override
    public int getOrder() {
        return 0;
    }
}

根据标识获取不同源
这里我们通过扩展AbstractRoutingDataSource来获取不同的源。它是Spring提供的一个可以根据用户发起的不同请求去转换不同的数据源，比如根据用户的不同地区语言选择不同的数据库。通过查看源码可以发现，它是通过determineCurrentLookupKey（）返回的不同key到sqlSessionFactory中获取不同源（前面已经展示了如何在sqlSessionFactory中注入多个源）

public class ReadWriteSplitRoutingDataSource extends AbstractRoutingDataSource {

    @Override
    protected Object determineCurrentLookupKey() {
        return DbContextHolder.getDbType();
    }
}

以上就完成了读写分离（多数据源）的配置方案。下面是一个具体的实例

使用方式
Entity

@Table(name = "t_sys_dic_type")
public class DicType extends BaseEntity{

    String code;

    String name;

    Integer status;

    ...
}

Mapper

public interface DicTypeMapper extends BaseMapper<DicType> {
}

Service

@Service
public class DicTypeService {
    @Autowired
    private DicTypeMapper dicTypeMapper;

    @ReadOnlyConnection
    public List<DicType> getAll(DicType dicType){
        if (dicType.getPage() != null && dicType.getRows() != null) {
            PageHelper.startPage(dicType.getPage(), dicType.getRows());
        }
        return dicTypeMapper.selectAll();
    }

}

注意这里的@ReadOnlyConnection注解

Controller

@RestController
@RequestMapping("/dictype")
public class DicTypeController {
    @Autowired
    private DicTypeService dicTypeService;

    @RequestMapping(value = "/all")
    public PageInfo<DicType> getALL(DicType dicType){
        List<DicType> dicTypeList = dicTypeService.getAll(dicType);
        return new PageInfo<>(dicTypeList);
    }
}

通过mvn spring-boot:run启动后

后台打印出

c.a.d.m.ReadOnlyConnectionInterceptor : set database connection to read only

说明使用了从库的链接获取数据

备注：如何保证多源事务呢？br/>1-在读写分离场景中不会考虑主从库事务，在纯读的上下文上使用@ReadOnlyConnection标签。其他则默认使用主库。
2-在多源场景中，Spring的@Transaction是可以保证多源的各自事务性的。