spring-boot 速成(9) druid+mybatis 多数据源及读写分离的处理

按上节继续学习，稍微复杂的业务系统，一般会将数据库按业务拆开，比如产品系统的数据库放在product db中，订单系统的数据库放在order db中...，然后，如果量大了，可能每个库还要考虑做读、写分离，以进一步提高系统性能，下面就来看看如何处理：

核心思路：配置多个数据源，然后利用RoutingDataSource结合AOP来动态切不同的库。

要解决的问题：

1、配置文件中，多数据源的配置节点如何设计？

druid:
    type: com.alibaba.druid.pool.DruidDataSource
    study:
      master: #study库的主库
        url: jdbc:mysql://localhost:3306/study?useSSL=false&characterEncoding=UTF-8&autoReconnect=true&zeroDateTimeBehavior=convertToNull&useUnicode=true
        driver-class-name: com.mysql.cj.jdbc.Driver
        username: root
        password: A1b2c3@def.com
        initial-size: 5
        min-idle: 1
        max-active: 20
        test-on-borrow: true
      slave: #study库的从库
        url: jdbc:mysql://localhost:3306/study_slave?useSSL=false&characterEncoding=UTF-8&autoReconnect=true&zeroDateTimeBehavior=convertToNull&useUnicode=true
        driver-class-name: com.mysql.cj.jdbc.Driver
        username: root
        password: A1b2c3@def.com
        initial-size: 5
        min-idle: 1
        max-active: 20
        test-on-borrow: true
    product:
      master: #product库的主库
        url: jdbc:mysql://localhost:3306/product?useSSL=false&characterEncoding=UTF-8&autoReconnect=true&zeroDateTimeBehavior=convertToNull&useUnicode=true
        driver-class-name: com.mysql.cj.jdbc.Driver
        username: root
        password: A1b2c3@def.com
        initial-size: 5
        min-idle: 1
        max-active: 20
        test-on-borrow: true
      slave: #product库的从库
        url: jdbc:mysql://localhost:3306/product_slave?useSSL=false&characterEncoding=UTF-8&autoReconnect=true&zeroDateTimeBehavior=convertToNull&useUnicode=true
        driver-class-name: com.mysql.cj.jdbc.Driver
        username: root
        password: A1b2c3@def.com
        initial-size: 5
        min-idle: 1
        max-active: 20
        test-on-borrow: true

上面的配置写法供参数，如果slave节点数要扩展，按这个格式，改造成slave1,slave2... 自行扩展。

2、配置类如何设计？

package com.cnblogs.yjmyzz.db.config;

/**
 * Created by jimmy on 6/18/17.
 */

import com.cnblogs.yjmyzz.db.datasource.DbContextHolder;
import com.cnblogs.yjmyzz.db.datasource.MasterSlaveRoutingDataSource;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.boot.autoconfigure.jdbc.DataSourceBuilder;
import org.springframework.boot.context.properties.ConfigurationProperties;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.Primary;
import org.springframework.jdbc.datasource.lookup.AbstractRoutingDataSource;
import org.springframework.transaction.annotation.EnableTransactionManagement;

import javax.sql.DataSource;
import java.util.HashMap;
import java.util.Map;


@Configuration
@EnableTransactionManagement
public class DataSourceConfiguration {

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

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

    @Bean(name = "studySlaveDataSource")
    @ConfigurationProperties(prefix = "druid.study.slave")
    public DataSource studySlaveDataSource1() {
        return DataSourceBuilder.create().type(dataSourceType).build();
    }

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

    @Bean(name = "productSlaveDataSource")
    @ConfigurationProperties(prefix = "druid.product.slave")
    public DataSource productSlaveDataSource1() {
        return DataSourceBuilder.create().type(dataSourceType).build();
    }

    @Bean(name = "dataSource")
    @Primary
    public AbstractRoutingDataSource dataSource() {
        MasterSlaveRoutingDataSource proxy = new MasterSlaveRoutingDataSource();
        Map<Object, Object> targetDataResources = new HashMap<>();
        targetDataResources.put(DbContextHolder.DbType.PRODUCT_MASTER, productMasterDataSource());
        targetDataResources.put(DbContextHolder.DbType.PRODUCT_SLAVE, productSlaveDataSource1());
        targetDataResources.put(DbContextHolder.DbType.STUDY_MASTER, studyMasterDataSource());
        targetDataResources.put(DbContextHolder.DbType.STUDY_SLAVE, studySlaveDataSource1());
        proxy.setDefaultTargetDataSource(productMasterDataSource());
        proxy.setTargetDataSources(targetDataResources);
        proxy.afterPropertiesSet();
        return proxy;
    }

}

参考这个，一看就明，不说多(注：@Primary一定要在动态数据源上，否则事务回滚无效！）

3、根据什么来切换db?

有很多选择，

a、用约定的方法前缀，比如：get/query/list开头的约定为读从库，其它为主库，但是这样还要考虑不同业务库的切换（即：何时切换到product库，何时切换到order库，可以再用不同的Scanner来处理，略复杂）

b、用自定义注解来处理，比如 @ProductMaster注解，表示切换到product的master库，这样同时把业务库，以及主还是从，一次性解决了，推荐这种。

这里，我定义了4个注解，代表product，study二个库的主及从。

4、aop在哪里拦截,如何拦截？

service层和mapper层都可以拦截，推荐在服务层拦截，否则如果一个业务方法里，即有读又有写，还得考虑如果遇到事务，要考虑的东西更多。

当然，如果拦截特定的注解，就不用过多考虑在哪个层，只认注解就行（当然，注解还是建议打在服务层上）。

dubbo-starter的一个小坑：spring boot中，只有managed bean才能用aop拦截，而dubbo-starter中的@service注解不是spring中的注解（是阿里package下的自定义注解），生成的service provider实例，aop拦截不到，解决办法，再加一个注解让spring认识它，参考：

Aop拦截类的参考代码如下：

package com.cnblogs.yjmyzz.db.aspect;

import com.cnblogs.yjmyzz.db.annotation.ProductMaster;
import com.cnblogs.yjmyzz.db.annotation.ProductSlave;
import com.cnblogs.yjmyzz.db.annotation.StudyMaster;
import com.cnblogs.yjmyzz.db.annotation.StudySlave;
import com.cnblogs.yjmyzz.db.datasource.DbContextHolder;
import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Aspect;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.core.Ordered;
import org.springframework.stereotype.Component;


@Aspect
@Component
public class MasterSlaveAspect implements Ordered {

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


    /**
     * 切换到product主库
     *
     * @param proceedingJoinPoint
     * @param productMaster
     * @return
     * @throws Throwable
     */
    @Around("@annotation(productMaster)")
    public Object proceed(ProceedingJoinPoint proceedingJoinPoint, ProductMaster productMaster) throws Throwable {
        try {
            logger.info("set database connection to product-master only");
            DbContextHolder.setDbType(DbContextHolder.DbType.PRODUCT_MASTER);
            Object result = proceedingJoinPoint.proceed();
            return result;
        } finally {
            DbContextHolder.clearDbType();
            logger.info("restore database connection");
        }
    }


    /**
     * 切换到product从库
     *
     * @param proceedingJoinPoint
     * @param productSlave
     * @return
     * @throws Throwable
     */
    @Around("@annotation(productSlave)")
    public Object proceed(ProceedingJoinPoint proceedingJoinPoint, ProductSlave productSlave) throws Throwable {
        try {
            logger.info("set database connection to product-slave only");
            DbContextHolder.setDbType(DbContextHolder.DbType.PRODUCT_SLAVE);
            Object result = proceedingJoinPoint.proceed();
            return result;
        } finally {
            DbContextHolder.clearDbType();
            logger.info("restore database connection");
        }
    }

    /**
     * 切换到study主库
     *
     * @param proceedingJoinPoint
     * @param studyMaster
     * @return
     * @throws Throwable
     */
    @Around("@annotation(studyMaster)")
    public Object proceed(ProceedingJoinPoint proceedingJoinPoint, StudyMaster studyMaster) throws Throwable {
        try {
            logger.info("set database connection to study-master only");
            DbContextHolder.setDbType(DbContextHolder.DbType.STUDY_MASTER);
            Object result = proceedingJoinPoint.proceed();
            return result;
        } finally {
            DbContextHolder.clearDbType();
            logger.info("restore database connection");
        }
    }

    /**
     * 切换到study从库
     *
     * @param proceedingJoinPoint
     * @param studySlave
     * @return
     * @throws Throwable
     */
    @Around("@annotation(studySlave)")
    public Object proceed(ProceedingJoinPoint proceedingJoinPoint, StudySlave studySlave) throws Throwable {
        try {
            logger.info("set database connection to study-slave only");
            DbContextHolder.setDbType(DbContextHolder.DbType.STUDY_SLAVE);
            Object result = proceedingJoinPoint.proceed();
            return result;
        } finally {
            DbContextHolder.clearDbType();
            logger.info("restore database connection");
        }
    }

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

5、其它事项

启用类上，一定要排除spring-boot自带的datasource配置，即：

@SpringBootApplication(exclude = {DataSourceAutoConfiguration.class})
@EnableAspectJAutoProxy
@ComponentScan("com.cnblogs.yjmyzz")
@MapperScan(basePackages = "com.cnblogs.yjmyzz.dao.mapper")
public class ServiceProvider {
    public static void main(String[] args) {
        SpringApplication.run(ServiceProvider.class, args);
    }
}

第1行：@SpringBootApplication(exclude = {DataSourceAutoConfiguration.class})

6、日志中如何输出格式化且带参数值的sql?

一般的sql输出是这样的：

我们可以把它变成下面这样：

是不是更友好！

方法：加一个mybtais的拦截器即可

package com.cnblogs.yjmyzz.db.interceptor;


import com.cnblogs.yjmyzz.util.PrettySQLFormatter;
import org.apache.ibatis.cache.CacheKey;
import org.apache.ibatis.executor.Executor;
import org.apache.ibatis.mapping.BoundSql;
import org.apache.ibatis.mapping.MappedStatement;
import org.apache.ibatis.mapping.ParameterMapping;
import org.apache.ibatis.plugin.*;
import org.apache.ibatis.reflection.MetaObject;
import org.apache.ibatis.session.Configuration;
import org.apache.ibatis.session.ResultHandler;
import org.apache.ibatis.session.RowBounds;
import org.apache.ibatis.type.TypeHandlerRegistry;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.text.SimpleDateFormat;
import java.util.Date;
import java.util.List;
import java.util.Properties;


/**
 * Created by 菩提树下的杨过(http://yjmyzz.cnblogs.com/) on 28/07/2017.
 */
@Intercepts({
        @Signature(type = Executor.class, method = "update", args = {MappedStatement.class, Object.class}),
        @Signature(type = Executor.class, method = "query", args = {MappedStatement.class, Object.class, RowBounds.class, ResultHandler.class}),
        @Signature(type = Executor.class, method = "query", args = {MappedStatement.class, Object.class, RowBounds.class, ResultHandler.class, CacheKey.class, BoundSql.class})})
public class MybatisInterceptor implements Interceptor {

    private static Logger logger = LoggerFactory.getLogger(MybatisInterceptor.class);

    private Properties properties;

    private final static SimpleDateFormat sdt = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");

    @Override
    public Object intercept(Invocation invocation) throws Throwable {
        MappedStatement mappedStatement = (MappedStatement) invocation.getArgs()[0];
        Object parameter = null;
        if (invocation.getArgs().length > 1) {
            parameter = invocation.getArgs()[1];
        }
        String sqlId = mappedStatement.getId();
        BoundSql boundSql = mappedStatement.getBoundSql(parameter);
        Configuration configuration = mappedStatement.getConfiguration();
        Object returnValue;
        long start = System.currentTimeMillis();
        returnValue = invocation.proceed();
        long end = System.currentTimeMillis();
        long time = (end - start);
        if (time > 1) {
            String sql = getSql(configuration, boundSql, sqlId, time);
            logger.debug("mapper method ==> " + sql.split("\^")[0] + "
，" + PrettySQLFormatter.getPrettySql(sql.split("\^")[1]) + "

，" + "sql execute time ==> " + time + " ms

");
        }
        return returnValue;
    }

    @Override
    public Object plugin(Object target) {
        return Plugin.wrap(target, this);
    }

    @Override
    public void setProperties(Properties properties) {
        this.properties = properties;
    }


    public static String getSql(Configuration configuration, BoundSql boundSql, String sqlId, long time) {
        String sql = showSql(configuration, boundSql);
        StringBuilder str = new StringBuilder(100);
        str.append(sqlId);
        str.append("^");
        str.append(sql);
        str.append("^");
        str.append(time);
        str.append("ms");
        return str.toString();
    }

    private static String getParameterValue(Object obj) {
        String value;
        if (obj instanceof String) {
            value = "'" + obj.toString() + "'";
        } else if (obj instanceof Date) {
            value = "'" + sdt.format(obj) + "'";
        } else {
            if (obj != null) {
                value = obj.toString();
            } else {
                value = "";
            }

        }
        return value;
    }

    public static String showSql(Configuration configuration, BoundSql boundSql) {
        Object parameterObject = boundSql.getParameterObject();
        List<ParameterMapping> parameterMappings = boundSql.getParameterMappings();
        String sql = boundSql.getSql().replaceAll("[\s]+", " ");
        if (parameterMappings.size() > 0 && parameterObject != null) {
            TypeHandlerRegistry typeHandlerRegistry = configuration.getTypeHandlerRegistry();
            if (typeHandlerRegistry.hasTypeHandler(parameterObject.getClass())) {
                sql = sql.replaceFirst("\?", getParameterValue(parameterObject));

            } else {
                MetaObject metaObject = configuration.newMetaObject(parameterObject);
                for (ParameterMapping parameterMapping : parameterMappings) {
                    String propertyName = parameterMapping.getProperty();
                    if (metaObject.hasGetter(propertyName)) {
                        Object obj = metaObject.getValue(propertyName);
                        sql = sql.replaceFirst("\?", getParameterValue(obj));
                    } else if (boundSql.hasAdditionalParameter(propertyName)) {
                        Object obj = boundSql.getAdditionalParameter(propertyName);
                        sql = sql.replaceFirst("\?", getParameterValue(obj));
                    }
                }
            }
        }
        return sql;
    }
}

这里面还用了hibernate的一个小工具，用于格式化sql

package com.cnblogs.yjmyzz.util;

import org.hibernate.engine.jdbc.internal.FormatStyle;

public class PrettySQLFormatter {

    public static void print(String sql) {
        System.out.println(FormatStyle.BASIC.getFormatter().format(sql));
    }


    public static void print(String remark, String sql) {
        System.out.println(remark
                + FormatStyle.BASIC.getFormatter().format(sql));
    }


    public static String getPrettySql(String sql) {
        return FormatStyle.BASIC.getFormatter().format(sql);
    }


    public static String getPrettySql(String remark, String sql) {
        return remark + FormatStyle.BASIC.getFormatter().format(sql);
    }

    public static void main(String[] args) {
        System.out.println(getPrettySql("select * from MyUser as A join MyFriend as B on A.id = B.pid where B.name like ? "));
    }
}

接下来，把这个拦截器配置在mybatis-config.xml里

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE configuration PUBLIC "-//mybatis.org//DTD Config 3.0//EN"
        "http://mybatis.org/dtd/mybatis-3-config.dtd">
<configuration>
    <settings>
        <setting name="cacheEnabled" value="true"/>
    </settings>

    <plugins>
        <plugin interceptor="com.cnblogs.yjmyzz.db.interceptor.MybatisInterceptor">
        </plugin>
    </plugins>

</configuration>

最后在application.yml里指定mybatis-config.xml所在的路径：

示例源码见：https://github.com/yjmyzz/spring-boot-dubbo-demo （dubbox2.8.5-multi-ds分支）