spring boot 连接多个数据源

在springboot中有需要连接多个数据源的情况。

首先配置文件application.properties中添加两个数据源连接字符串

mybatis.type-aliases-package=com.murphy.boot.model


spring.datasource.source1.driver-class-name = oracle.jdbc.driver.OracleDriver
spring.datasource.source1.url = jdbc:oracle:thin:@localhost:1521/ORCL
spring.datasource.source1.username = DCMSYA
spring.datasource.source1.password = DCMS
#使用Druid数据源
spring.datasource.source1.initialSize=5
# 初始化大小，最小，最大
spring.datasource.source1.minIdle=5
spring.datasource.source1.maxActive= 20
# 配置获取连接等待超时的时间
spring.datasource.source1.maxWait= 60000
# 配置间隔多久才进行一次检测，检测需要关闭的空闲连接，单位是毫秒
spring.datasource.source1.timeBetweenEvictionRunsMillis= 60000
# 配置一个连接在池中最小生存的时间，单位是毫秒
spring.datasource.source1.minEvictableIdleTimeMillis= 300000
spring.datasource.source1.validationQuery= select 1 from dual
spring.datasource.source1.testWhileIdle= true
spring.datasource.source1.testOnBorrow= false
spring.datasource.source1.testOnReturn= false
# 打开PSCache，并且指定每个连接上PSCache的大小
spring.datasource.source1.poolPreparedStatements= true
spring.datasource.source1.maxPoolPreparedStatementPerConnectionSize= 20
# 配置监控统计拦截的filters，去掉后监控界面sql无法统计，'wall'用于防火墙
spring.datasource.source1.filters= stat,slf4j


spring.datasource.source2.driver-class-name = oracle.jdbc.driver.OracleDriver
spring.datasource.source2.url = jdbc:oracle:thin:@localhost:1521/ORCL
spring.datasource.source2.username = DCMSYA_LIVE
spring.datasource.source2.password = DCMS
#使用Druid数据源
spring.datasource.source2.initialSize=5
# 初始化大小，最小，最大
spring.datasource.source2.minIdle=5
spring.datasource.source2.maxActive= 20
# 配置获取连接等待超时的时间
spring.datasource.source2.maxWait= 60000
# 配置间隔多久才进行一次检测，检测需要关闭的空闲连接，单位是毫秒
spring.datasource.source2.timeBetweenEvictionRunsMillis= 60000
# 配置一个连接在池中最小生存的时间，单位是毫秒
spring.datasource.source2.minEvictableIdleTimeMillis= 300000
spring.datasource.source2.validationQuery= select 1 from dual
spring.datasource.source2.testWhileIdle= true
spring.datasource.source2.testOnBorrow= false
spring.datasource.source2.testOnReturn= false
# 打开PSCache，并且指定每个连接上PSCache的大小
spring.datasource.source2.poolPreparedStatements= true
spring.datasource.source2.maxPoolPreparedStatementPerConnectionSize= 20
# 配置监控统计拦截的filters，去掉后监控界面sql无法统计，'wall'用于防火墙
spring.datasource.source2.filters= stat,slf4j

添加数据库资源类DataSourceOneConfig

@Configuration
//扫描Mapper  basePackages要精确到source1目录便于进行不同数据源的区分
@MapperScan(basePackages = "com.murphy.boot.mapper.source1", sqlSessionTemplateRef = "sqlSessionTemplateOne")
public class DataSourceOneConfig {

    @Bean(name = "dataSourceOne")
    @ConfigurationProperties(prefix = "spring.datasource.source1")
    @Primary //设置主数据源
    public DataSource DataSourceOne(){
        DruidDataSource dataSource = new DruidDataSource();
        return dataSource;

    }

    @Bean
    public ServletRegistrationBean druidStatViewServlet() {
        ServletRegistrationBean registrationBean = new ServletRegistrationBean(new StatViewServlet(), "/druid/*");
        registrationBean.addInitParameter("allow", "127.0.0.1"); // IP白名单 (没有配置或者为空，则允许所有访问)
        registrationBean.addInitParameter("deny", ""); // IP黑名单 (存在共同时，deny优先于allow)
        registrationBean.addInitParameter("loginUsername", "admin");
        registrationBean.addInitParameter("loginPassword", "admin");
        registrationBean.addInitParameter("resetEnable", "false");
        return registrationBean;
    }

    @Bean
    public FilterRegistrationBean druidWebStatViewFilter() {
        FilterRegistrationBean registrationBean = new FilterRegistrationBean(new WebStatFilter());
        registrationBean.addInitParameter("urlPatterns", "/*");
        registrationBean.addInitParameter("exclusions", "*.js,*.gif,*.jpg,*.bmp,*.png,*.css,*.ico,/druid/*");
        return registrationBean;
    }


    @Bean(name = "sqlSessionFactoryOne")
    @Primary
    public SqlSessionFactory sqlSessionFactoryOne(@Qualifier("dataSourceOne") DataSource dataSource)throws Exception{
        SqlSessionFactoryBean bean = new SqlSessionFactoryBean();
        bean.setDataSource(dataSource);
        return bean.getObject();
    }

    @Bean(name = "dataSourceTransactionManagerOne")
    @Primary
    public DataSourceTransactionManager dataSourceTransactionManagerOne(@Qualifier("dataSourceOne") DataSource dataSource){
        return new DataSourceTransactionManager(dataSource);
    }

    @Bean(name = "sqlSessionTemplateOne")
    @Primary
    public SqlSessionTemplate sqlSessionTemplateOne(@Qualifier("sqlSessionFactoryOne") SqlSessionFactory sqlSessionFactory)throws Exception{
        return new SqlSessionTemplate(sqlSessionFactory);
    }


}

本人用的是 tk.mybatis和 druid连接池 jar包：

<!-- druid连接池-->
        <dependency>
            <groupId>com.alibaba</groupId>
            <artifactId>druid</artifactId>
            <version>1.1.8</version>
        </dependency>
        <!--tk通用mapper-->
        <dependency>
            <groupId>tk.mybatis</groupId>
            <artifactId>mapper-spring-boot-starter</artifactId>
            <version>1.2.4</version>
        </dependency>

同样的，创建DataSourceTwoConifg,设置配置文件为 spring.datasource.source2，并指定对应的mapper文件位置为source2,。

创建Mapper文件，先创建父接口CommnMapper

/**
 * @Author Murphy
 * @Date 2018-10-18 14:50
 */
public interface CommnMapper<T> extends Mapper<T>, MySqlMapper<T> {

}

在mapper文件夹下创建source1和source2，即两个数据源类中配置的地址。

分别创建对应的mapper，集成CommnMapper

/**
 * @Author Murphy
 * @Date 2018-10-18 14:53
 */
@Component
@Mapper
public interface CaseDeptDealOneMapper extends CommnMapper<CaseDeptDeal> {

}

分别访问两个文件夹下的mapper就可以访问两个数据库了。

目录结构为：

 

 

测试代码

    @Autowired
    private CaseModelOneMapper userInfoOneMapper;
    @Autowired
    private CaseModelTwoMapper userInfoTwoMapper;
    @Test
    public void getUer(){
        List<Casehistory> caseModels = userInfoTwoMapper.selectAll();
        for (int i = 0; i < caseModels.size(); i++) {
            System.out.println(caseModels.get(i).getCaseid());

        }
    }