springboot使用多数据源以及配置

1. 首先在application中配置数据源地址

my.datasource.koi.type=com.alibaba.druid.pool.DruidDataSource
my.datasource.koi.url=jdbc:mysql://localhost:3306/?useUnicode=true&characterEncoding=utf-8&useSSL=true
my.datasource.koi.username=root
my.datasource.koi.password=123456
my.datasource.koi.driverClassName=com.mysql.jdbc.Driver

　　

2. 在@Bean中 new DataSource的地址空间

@Configuration
public class KOIDataSourceConfig {

    @Resource
    private DataSourceFactory dataSourceFactory;

    @Bean(name = "koiDataSource")
    @ConfigurationProperties(prefix = "my.datasource.koi")
    public DruidDataSource createDataSource() {
        return dataSourceFactory.create();
    }
    // dataSourceFactory.create() : DruidDataSource dataSource = new DruidDataSource();

    @Bean(name = "jdbcTemplate")
    public JdbcTemplate createSqlSessionTemplate(@Qualifier("koiDataSource") DruidDataSource druidDataSource) {
        return new JdbcTemplate(druidDataSource);
    }
}

　　

3.  在Service中直接使用即可

@Service
public class KOISchedulerImpl implements KOIScheduler{

    @Resource
    private JdbcTemplate jdbcTemplate;

    @Transactional
    @Override
    public void set_task_and_run(SQLPool pool, String config_path) {

        ExecuteSQL.prepare_and_execut_sql(pool, config_path);

    }

    @Transactional
    @Override
    public void startSchedulerTaskByDruid(String configPath) {
        jdbcTemplate.execute("INSERT INTO mydata.`test` (id) VALUE(1);");
    }
}

　　

读取application中的配置信息:

1.  在application中配置  spring.sql.path = E://z

2.  定义一个Component类

@Component
public class QualityRecode {

    @Value("${spring.sql.path}")
    private String sqlPath;

    public QualityRecode() {}

    public String getSqlPath() {
        return sqlPath;
    }

    public void setSqlPath(String sqlPath) {
        this.sqlPath = sqlPath;
    }
}

3. 在Controller中实例化该类， 获取数据

@Resource
private QualityRecode recode;

recode.getSqlPath()

　　

4.  使用Mapper读取 写入数据

1.  先在application配置数据源信息

2.  配置数据源config类

@Configuration
@MapperScan(basePackages = "com.nio.mapper.dcs", sqlSessionTemplateRef = "dcsSqlSessionTemplate")
public class DcsDataSourceConfig {

    @Resource
    private DataSourceFactory dataSourceFactory;

    @Bean(name = "dcsDataSource")
    @ConfigurationProperties(prefix = "my.datasource.dcs")
    public DruidDataSource createDataSource() {
        return dataSourceFactory.create();
    }

    @Bean(name = "dcsSqlSessionFactory")
    public SqlSessionFactory createSqlSessionFactory(@Qualifier("dcsDataSource") DataSource dataSource) throws Exception {
        SqlSessionFactoryBean bean = new SqlSessionFactoryBean();
        bean.setDataSource(dataSource);
        bean.setMapperLocations(
                new PathMatchingResourcePatternResolver().getResources("classpath:mybatis/mapper/dcs/*.xml"));
        return bean.getObject();
    }

//    @Bean(name = "cmsTransactionManager")
//    public DataSourceTransactionManager createTransactionManager(@Qualifier("cmsDataSource") DataSource dataSource) {
//        return new DataSourceTransactionManager(dataSource);
//    }

    @Bean(name = "dcsSqlSessionTemplate")
    public SqlSessionTemplate createSqlSessionTemplate(@Qualifier("dcsSqlSessionFactory") SqlSessionFactory sqlSessionFactory) {
        return new SqlSessionTemplate(sqlSessionFactory);
    }
}

3.  创建实体类 接收表传来的数据

4.  编辑Mapper类， 创建查询-插入方法

public interface CmsMapper {
    List<PostEntity> getPostEntityList();
    List<PostEntity> getLatestPostEntityList(@Param("yesterdayTimeStamp") Long yesterdayTimeStamp,
                                             @Param("currentTimeStamp") Long currentTimeStamp);

void insertError(@Param("TaskLogRecordEntity") List<TaskLogRecord> taskLogRecords);
}

5.  实现Mapper类

<select id="getPostEntityList" resultType="com.nio.entities.PostEntity">
        SELECT
            up.account_id AS accountId,
            up.create_time AS updateTime,
            up.content,
            up.like_count AS likeCount,
            up.comment_count AS commentCount,
            lu.`name` AS nickName
        FROM
            lifestyle_prod.user_post up
                LEFT JOIN lifestyle_prod.livestream_user lu ON up.account_id = lu.user_id
        WHERE
            up.STATUS = 1
          AND up.resource_type = 'user_post'
          AND lu.supplier = 'rongyun'
          AND up.create_time >= #{yesterdayTimeStamp}
          AND up.create_time <![CDATA[<=]]> #{currentTimeStamp}
    </select>

    <insert id="insertError">
        insert into `task_log_record`
        (
        id,task_type,execute_time,error_info
        )
        values
        <foreach collection="TaskLogRecordEntity" item="error" separator=",">
            (
            #{error.id},
            #{error.task_type},
            #{error.execute_time},
            #{error.error_info}
            )
        </foreach>
    </insert>

　　

6.  需要注意的问题:

1.  当查询一个原表中不存在，  但需要写在中间实体的，  指标查询表时：

例如， MySQL   该表只有一个字段 id ,   在计算之后， 生成

public class TestEntity {

    private long id2;

    private long id3;

}

Mapper应写成:  

    <select id="getTest" resultType="com.nio.entities.TestEntity">

    select t.id + 3 id2, t.id +1 id3 from
    (select 2 id from test) t
    </select>

## select 的值， 必须与实体类名一致， 而不是表中字段的名;