[大数据]Mybatis缓存机制

Mybatis 提供一级缓存和二级缓存的机制。
一级缓存是 SQLSession 级别的缓存。在操作数据库时，每个 SqlSession 类的实例对象中有一个数据结构（HashMap）可以用于存储缓存数据。不同的 SqlSession 类的实例对象的数据区域（HashMap）是互不影响的。
二级缓存是 Mapper 级别的缓存，多个 SqlSession 类的实例对象操作同一个 Mapper 配置文件中的 SQL 语句，多个 SqlSession 类的实例对象可以共用二级缓存，二级缓存是跨 SqlSession 的。
Mybatis 的缓存模式如下图所示：

每个 SqlSession 类的实例对象自身有一个一级缓存，而查询同一个 Mapper 映射文件的 SqlSession 类的实例对象之间又共享同一个二级缓存。

一、一级查询缓存

一级查询缓存存在于每一个 SqlSession 类的实例对象中，当第一次查询某一个数据时，SqlSession 类的实例对象会将该数据存入一级缓存区域，在没有收到改变数据的请求之前，用户再次查询该数据，都会从缓存中获取该数据，而不是再次连接数据库进行查询。

1.1 一级缓存原理阐述

上图阐述了一个 SqlSession 类的实例对象下的一级缓存的工作原理。当第一次查询 id 为 1 的用户信息时，SqlSession 首先到一级缓存区域查询，如果没有相关数据，则从数据库查询。然后 SqlSession 将查询结果保存到一级缓存区域。在下一次查询时，如果 SqlSession 执行了 commit 操作（即执行了修改、添加和删除），则会清空它的一级缓存区域，以此来保证缓存中的信息是最新的，避免脏读现象发生。如果在这期间 SqlSession 一直没有执行 commit 操作修改数据，那么下一次查询 id 为 1 的用户信息时，SqlSession 在一级缓存中就会发现该信息，然后从缓存中获取用户信息。

1.2 一级缓存测试示例

Mybatis 默认支持一级缓存，不需要在配置文件中配置一级缓存的相关数据。
使用同一个 SqlSession 对象，对 id 为 1 的员工查询两次。

EmployeeCacheMapper.xml

<?xml version="1.0" encoding="UTF-8" ?>
<!DOCTYPE mapper
        PUBLIC "-//mybatis.org//DTD Mapper 3.0//EN"
        "http://mybatis.org/dtd/mybatis-3-mapper.dtd">
<mapper namespace="com.example.mapper.EmployeeCacheMapper">

    <select id="getEmpById" resultType="com.example.pojo.Employee">
        select * from tbl_employee where id = #{id}
    </select>

</mapper>

EmployeeCacheMapper.java

package com.example.mapper;

import com.example.pojo.Employee;

public interface EmployeeCacheMapper {

    public Employee getEmpById(Integer id);
}

测试方法：

package com.example;

import com.example.mapper.EmployeeCacheMapper;
import com.example.pojo.Employee;
import org.apache.ibatis.io.Resources;
import org.apache.ibatis.session.SqlSession;
import org.apache.ibatis.session.SqlSessionFactory;
import org.apache.ibatis.session.SqlSessionFactoryBuilder;
import org.testng.annotations.Test;

import java.io.IOException;
import java.io.InputStream;

public class MybatisCacheTest {

    public SqlSessionFactory getSqlSessionFactory() throws IOException {
        InputStream inputStream = Resources.getResourceAsStream("mybatis-config.xml");
        SqlSessionFactory sqlSessionFactory = new SqlSessionFactoryBuilder().build(inputStream);
        return sqlSessionFactory;
    }

    @Test
    public void test01() throws IOException {
        SqlSessionFactory sqlSessionFactory = getSqlSessionFactory();
        SqlSession sqlSession = sqlSessionFactory.openSession();
        try{
            EmployeeCacheMapper mapper = sqlSession.getMapper(EmployeeCacheMapper.class);
            Employee emp01 = mapper.getEmpById(1);
            System.out.println(emp01);
            Employee emp02 = mapper.getEmpById(1);
            System.out.println(emp02);
            System.out.println(emp01 == emp02);
        } finally {
            sqlSession.close();
        }
    }
}

控制台结果：

[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
true

当第一次查询 id 为 1 的员工信息时，执行了 SQL 语句，而第二次查询 id 为 1 的员工信息时，没有任何的日志输出。而对于两次取出的员工对象的比对，也证明了第二次数据不是从数据库查询出来的，是从一级缓存中获取的。

1.3 一级缓存失效的四种情况

1、SqlSession 不同

测试语句如下：

@Test
public void test02() throws IOException {
    SqlSessionFactory sqlSessionFactory = getSqlSessionFactory();
    SqlSession sqlSession = sqlSessionFactory.openSession();
    try{
        EmployeeCacheMapper mapper = sqlSession.getMapper(EmployeeCacheMapper.class);
        Employee emp01 = mapper.getEmpById(1);
        System.out.println(emp01);
        SqlSession sqlSession1 = sqlSessionFactory.openSession();
        EmployeeCacheMapper mapper1 = sqlSession1.getMapper(EmployeeCacheMapper.class);
        Employee emp02 = mapper1.getEmpById(1);
        System.out.println(emp02);
        System.out.println(emp01 == emp02);
    } finally {
        sqlSession.close();
    }
}

控制台结果：

[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}

[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
false

2、SqlSession 相同，查询条件不同

分别查询 id 为 1 和 id 为 2 个员工信息
测试方法：

 @Test
 public void test03() throws IOException {
     SqlSessionFactory sqlSessionFactory = getSqlSessionFactory();
     SqlSession sqlSession = sqlSessionFactory.openSession();
     try{
         EmployeeCacheMapper mapper = sqlSession.getMapper(EmployeeCacheMapper.class);
         Employee emp01 = mapper.getEmpById(1);
         System.out.println(emp01);
         Employee emp02 = mapper.getEmpById(2);
         System.out.println(emp02);
         System.out.println(emp01 == emp02);
     } finally {
         sqlSession.close();
     }
 }

控制台结果：

[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 2(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=2, lastName='jerry', email='jerry@123.com', gender='0', department=null}
false

3、SqlSession 相同，两次查询之间执行了增删改操作

增删改操作可能会对数据有影响
EmployeeCacheMapper.xml：

<insert id="addEmp" databaseId="mysql" parameterType="employee" useGeneratedKeys="true" keyProperty="id">
    insert into tbl_employee (id, last_name, email, gender)
    values (#{id}, #{lastName}, #{email}, #{gender})
</insert>

EmployeeCacheMapper.java

package com.example.mapper;

import com.example.pojo.Employee;

public interface EmployeeCacheMapper {
    public void addEmp(Employee employee);
}

测试方法：

@Test
public void test04() throws IOException {
    SqlSessionFactory sqlSessionFactory = getSqlSessionFactory();
    SqlSession sqlSession = sqlSessionFactory.openSession();
    try{
        EmployeeCacheMapper mapper = sqlSession.getMapper(EmployeeCacheMapper.class);
        Employee emp01 = mapper.getEmpById(1);
        System.out.println(emp01);
        Employee employee = new Employee();
        employee.setLastName("meichaofeng");
        employee.setEmail("meichaofeng@123.com");
        employee.setGender("0");
        mapper.addEmp(employee);
        Employee emp02 = mapper.getEmpById(1);
        System.out.println(emp02);
        System.out.println(emp01 == emp02);
    } finally {
        sqlSession.close();
    }
}

控制台结果：

[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
[main] [com.example.mapper.EmployeeCacheMapper.addEmp]-[DEBUG] ==>  Preparing: insert into tbl_employee (id, last_name, email, gender) values (?, ?, ?, ?) 
[main] [com.example.mapper.EmployeeCacheMapper.addEmp]-[DEBUG] ==> Parameters: null, meichaofeng(String), meichaofeng@123.com(String), 0(String)
[main] [com.example.mapper.EmployeeCacheMapper.addEmp]-[DEBUG] <==    Updates: 1
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
false

4、SqlSession 相同，但手动清除了一级缓存

测试方法：

 @Test
 public void test05() throws IOException {
     SqlSessionFactory sqlSessionFactory = getSqlSessionFactory();
     SqlSession sqlSession = sqlSessionFactory.openSession();
     try{
         EmployeeCacheMapper mapper = sqlSession.getMapper(EmployeeCacheMapper.class);
         Employee emp01 = mapper.getEmpById(1);
         System.out.println(emp01);
         sqlSession.clearCache(); // 清空一级缓存
         Employee emp02 = mapper.getEmpById(1);
         System.out.println(emp02);
         System.out.println(emp01 == emp02);
     } finally {
         sqlSession.close();
     }
 }

控制台结果：

[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==>  Preparing: select * from tbl_employee where id = ? 
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] ==> Parameters: 1(Integer)
[main] [com.example.mapper.EmployeeCacheMapper.getEmpById]-[DEBUG] <==      Total: 1
Employee{id=1, lastName='tomcat', email='tom@123.com', gender='0', department=null}
false