Spring Boot最大的特点是简化开发，因此使用Java Config实现去xml配置，本文将使用这种方式完成对阿里云ADS和RDS两个产品的多数据源配置。同时，会用到阿里巴巴的开源数据源Druid。

依赖

• 因为ADS和RDS都兼容mysql的JDBC实现，因此工程只需要依赖mysql的driver。
• Spring Boot易用性的特点是通过Starter实现的，我们只需依赖mybatis的spring-boot-starter即可。
• Druid是数据源。

<dependency>
    <groupId>org.mybatis.spring.boot</groupId>
    <artifactId>mybatis-spring-boot-starter</artifactId>
    <version>${mybatis-spring-boot-starter.version}</version>
</dependency>
<dependency>
    <groupId>com.alibaba</groupId>
    <artifactId>druid</artifactId>
    <version>${druid.version}</version>
</dependency>
<dependency>
    <groupId>mysql</groupId>
    <artifactId>mysql-connector-java</artifactId>
    <version>${mysql-connector-java.version}</version>
</dependency>
<properties>
    <mysql-connector-java.version>5.1.34</mysql-connector-java.version>
    <druid.version>1.0.18</druid.version>
    <mybatis-spring-boot-starter.version>1.1.1</mybatis-spring-boot-starter.version>
</properties>

数据源配置

多数据源配置要求必须有一个是『主』，我们将RDS设定为主，配置如下：

@Configuration
@MapperScan(basePackages = RdsDataSourceConfig.PACKAGE, sqlSessionFactoryRef = "rdsSqlSessionFactory")
public class RdsDataSourceConfig {
    static final String PACKAGE = "com.aliyun.xxx.repository.mybatis.rds";

    @Value("${xxx_mysql_url}")
    private String dbUrl;
    @Value("${xxx_mysql_username}")
    private String dbUser;
    @Value("${xxx_mysql_password}")
    private String dbPassword;

    @Bean(name = "rdsDataSource")
    @Primary
    public DataSource rdsDataSource() {
        DruidDataSource dataSource = new DruidDataSource();
        dataSource.setDriverClassName("com.mysql.jdbc.Driver");
        dataSource.setUrl(dbUrl);
        dataSource.setUsername(dbUser);
        dataSource.setPassword(dbPassword);
        return dataSource;
    }

    @Bean(name = "rdsTransactionManager")
    @Primary
    public DataSourceTransactionManager rdsTransactionManager() {
        return new DataSourceTransactionManager(rdsDataSource());
    }

    @Bean(name = "rdsSqlSessionFactory")
    @Primary
    public SqlSessionFactory rdsSqlSessionFactory(@Qualifier("rdsDataSource") DataSource rdsDataSource) throws Exception {
        final SqlSessionFactoryBean sessionFactory = new SqlSessionFactoryBean();
        sessionFactory.setDataSource(rdsDataSource);
        return sessionFactory.getObject();
    }
}

• @Value将获取Spring Boot全局配置文件的kv配置。也可以通过Environment对象的getProperty方法获得kv配置。
• @MapperScan配置要扫描并容器管理的包路径，这里我们精确到rds，以便使其内的Mapper（见下文）与其他数据源隔离。
• sqlSessionFactoryRef要定义key唯一的SqlSessionFactory实例，否则同名的处理是后者被忽略。

类似地，另一个数据源ADS的配置如下：

@Configuration
@MapperScan(basePackages = AdsDataSourceConfig.PACKAGE, sqlSessionFactoryRef = "adsSqlSessionFactory")
public class AdsDataSourceConfig {
    static final String PACKAGE = "com.aliyun.xxx.repository.mybatis.ads";

    @Value("${xxx_ads_url}")
    private String dbUrl;
    @Value("${xxx_ads_username}")
    private String dbUser;
    @Value("${xxx_ads_password}")
    private String dbPassword;

    @Bean(name = "adsDataSource")
    public DataSource adsDataSource() {
        DruidDataSource dataSource = new DruidDataSource();
        dataSource.setDriverClassName("com.mysql.jdbc.Driver");
        dataSource.setUrl(dbUrl);
        dataSource.setUsername(dbUser);
        dataSource.setPassword(dbPassword);
        return dataSource;
    }

    @Bean(name = "adsTransactionManager")
    public DataSourceTransactionManager adsTransactionManager(@Qualifier("adsDataSource") DataSource adsDataSource) {
        return new DataSourceTransactionManager(adsDataSource);
    }

    @Bean(name = "adsSqlSessionFactory")
    public SqlSessionFactory adsSqlSessionFactory(@Qualifier("adsDataSource") DataSource adsDataSource) throws Exception {
        final SqlSessionFactoryBean sessionFactory = new SqlSessionFactoryBean();
        sessionFactory.setDataSource(adsDataSource);
        return sessionFactory.getObject();
    }
}

Mapper

Mapper是mybatis的DAL层接口，其实现是由mybatis内部完成的，也就是说，我们只需要定义好接口即可完成CRUD。

com.aliyun.xxx.repository.mybatis.rds包内的Mapper只会用到RDS数据源。

package com.aliyun.xxx.repository.mybatis.rds;

@Mapper
public interface MetaMapper {
    @SelectProvider(type = MetaSelectProvider.class, method = "xxxQueryById")
    @Results(value ={
            @Result(id=true, property="id",column="id",javaType=Long.class,jdbcType=JdbcType.BIGINT),
            ...
            @Result(property="gmtCreate",column="gmt_create",javaType=Date.class,jdbcType=JdbcType.TIMESTAMP)})
    List<XxxPojo> xxxQueryById(@Param("xxxId") String xxxId);
}

上述Mapper定义了一个Select方法，@SelectProvider用于定义具体的SQL语句，@Results用于定义具体的表字段与POJO属性的映射。

MetaSelectProvider可以使用MyBatis的API完成SQL的拼装。

public class MetaSelectProvider {
    private static final String xxxSQL = "id, ...,gmt_create";

    public String xxxQueryById(Map<String, Object> parameter) {
        org.apache.ibatis.jdbc.SQL sql = new org.apache.ibatis.jdbc.SQL();
        sql.SELECT(xxxSQL).FROM("table_name").WHERE("biz_id=#{xxxId}");
        return sql.toString();
    }
}

DAO

@Repository
public class XxxDAO {
    @Autowired
    private MetaMapper metaMapper;

    public List<XxxPojo> xxxQueryById(String xxxId) throws DataAccessException {
        return metaMapper.xxxQueryById(xxxId);
    }
}

SpringBootApplication

Spring Boot的入口类定义如下：

@SpringBootApplication
@PropertySource("file:/opt/xxx/xxx.properties")
public class XxxApplication {
    public static void main(String[] args) {
        SpringApplication.run(XxxApplication.class, args);
    }
}

测试类

使用Spring的上下文，完成单元测试：

@RunWith(SpringJUnit4ClassRunner.class)
@SpringApplicationConfiguration(classes = XxxApplication.class)
public class TestXxxDAO {
    private static final Logger logger = LogManager.getLogger(RelevanceResource.class);
    @Autowired
    private XxxDAO dao;

    @Test
    public void testXxxById() {
        String xxxId = "C00018";
        List<XxxPojo> list = dao.xxxQueryById(xxxId);
        Assert.assertNotNull(list);
        logger.info(...);
    }
}