Mybatis缓存介绍及原理分析

一. ⼀级缓存

（一）在⼀个sqlSession中，对User表根据id进⾏两次查询，查看他们发出sql语句的情况

@Test
    public void test1(){
//根据 sqlSessionFactory 产⽣ session
        SqlSession sqlSession = sessionFactory.openSession();
        UserMapper userMapper = sqlSession.getMapper(UserMapper.class);
        //第⼀次查询，发出sql语句，并将查询出来的结果放进缓存中
        User u1 = userMapper.selectUserByUserId(1);
        System.out.println(u1);
        //第⼆次查询，由于是同⼀个sqlSession,会在缓存中查询结果
        //如果有，则直接从缓存中取出来，不和数据库进⾏交互
        User u2 = userMapper.selectUserByUserId(1);
        System.out.println(u2);
        sqlSession.close();
    }

查看控制台打印情况：

（二）同样是对user表进⾏两次查询，只不过两次查询之间进⾏了⼀次update操作。

@Test
    public void test2(){
        //根据 sqlSessionFactory 产⽣ session
        SqlSession sqlSession = sessionFactory.openSession();
        UserMapper userMapper = sqlSession.getMapper(UserMapper.class);
        //第⼀次查询，发出sql语句，并将查询的结果放⼊缓存中
        User u1 = userMapper.selectUserByUserId( 1 );
        System.out.println(u1);
        //第⼆步进⾏了⼀次更新操作，sqlSession.commit()
        u1.setSex("⼥");
        userMapper.updateUserByUserId(u1);
        sqlSession.commit();
        //第⼆次查询，由于是同⼀个sqlSession.commit(),会清空缓存信息
        //则此次查询也会发出sql语句
        User u2 = userMapper.selectUserByUserId(1);
        System.out.println(u2);
        sqlSession.close();
    }

查看控制台打印情况：

 （三）总结

1. 第⼀次发起查询⽤户id为1的⽤户信息，先去找缓存中是否有id为1的⽤户信息，如果没有，从 数据库查询⽤户 信息。得到⽤户信息，将⽤户信息存储到⼀级缓存中。

2. 如果中间sqlSession去执⾏commit操作（执⾏插⼊、更新、删除），则会清空SqlSession中的 ⼀级缓存，这 样做的⽬的为了让缓存中存储的是最新的信息，避免脏读。

3. 第⼆次发起查询⽤户id为1的⽤户信息，先去找缓存中是否有id为1的⽤户信息，缓存中有，直 接从缓存中获取 ⽤户信息

 ⼀级缓存原理探究与源码分析

CacheKey cacheKey = new CacheKey();
//MappedStatement 的 id
// id就是Sql语句的所在位置包名+类名+ SQL名称
cacheKey.update(ms.getId());
// offset 就是 0
cacheKey.update(rowBounds.getOffset());
// limit 就是 Integer.MAXVALUE
cacheKey.update(rowBounds.getLimit());
//具体的SQL语句
cacheKey.update(boundSql.getSql());
//后⾯是update 了 sql中带的参数
cacheKey.update(value);
...
if (configuration.getEnvironment() != null) {
// issue #176
cacheKey.update(configuration.getEnvironment().getId());
}

创建缓存key会经过⼀系列的update⽅法，udate⽅法由⼀个CacheKey这个对象来执⾏的，这个 update⽅法最终 由updateList的list来把五个值存进去，对照上⾯的代码和下⾯的图示，你应该能 理解这五个值都是什么了

 

这⾥需要注意⼀下最后⼀个值，configuration.getEnvironment().getId()这是什么，这其实就是 定义在mybatisconfig.xml中的标签，⻅如下：

<environments default="development"> <environment id="development"> <transactionManager type="JDBC"/>
<dataSource type="POOLED">
 <property name="driver" value="${jdbc.driver}"/>
 <property name="url" value="${jdbc.url}"/>
 <property name="username" value="${jdbc.username}"/>
 <property name="password" value="${jdbc.password}"/>
 </dataSource>
</environment>
</environments>

那么我们回归正题，那么创建完缓存之后该⽤在何处呢？总不会凭空创建⼀个缓存不使⽤吧？绝对不会 的，经过我 们对⼀级缓存的探究之后，我们发现⼀级缓存更多是⽤于查询操作，毕竟⼀级缓存也叫做查 询缓存吧，为什么叫查 询缓存我们⼀会⼉说。我们先来看⼀下这个缓存到底⽤在哪了，我们跟踪到 query⽅法如下：

Override
public <E> List<E> query(MappedStatement ms, Object parameter, RowBounds rowBounds,
        ResultHandler resultHandler) throws SQLException {
        BoundSql boundSql = ms.getBoundSql(parameter);
        //创建缓存
        CacheKey key = createCacheKey(ms, parameter, rowBounds, boundSql);
        return query(ms, parameter, rowBounds, resultHandler, key, boundSql);
        }
@SuppressWarnings("unchecked")
Override
public <E> List<E> query(MappedStatement ms, Object parameter, RowBounds rowBounds,
        ResultHandler resultHandler, CacheKey key, BoundSql boundSql) throws SQLException {
        ...
        list = resultHandler == null ? (List<E>) localCache.getObject(key) : null; if (list
        != null) {
        //这个主要是处理存储过程⽤的。
        handleLocallyCachedOutputParameters(ms, key, parameter, boundSql);
        } else {
        list = queryFromDatabase(ms, parameter, rowBounds, resultHandler, key, boundSql);
        }
        ...
        }
// queryFromDatabase ⽅法
private <E> List<E> queryFromDatabase(MappedStatement ms, Object parameter, RowBounds
        rowBounds, ResultHandler resultHandler, CacheKey key, BoundSql boundSql) throws
        SQLException {
        List<E> list;
        localCache.putObject(key, EXECUTION_PLACEHOLDER);
        try {
        list = doQuery(ms, parameter, rowBounds, resultHandler, boundSql);
        } finally {
        localCache.removeObject(key);
        }
        localCache.putObject(key, list);
        if (ms.getStatementType() == StatementType.CALLABLE) {
        localOutputParameterCache.putObject(key, parameter);
        }
        return list; 
}

如果查不到的话，就从数据库查，在queryFromDatabase中，会对localcache进⾏写⼊。 localcache对象的put⽅ 法最终交给Map进⾏存放

private Map<Object, Object> cache = new HashMap<Object, Object>();
 @Override
 public void putObject(Object key, Object value) { cache.put(key, value);
}

二. ⼆级缓存

⼆级缓存的原理和⼀级缓存原理⼀样，第⼀次查询，会将数据放⼊缓存中，然后第⼆次查询则会直接去 缓存中取。 但是⼀级缓存是基于sqlSession的，⽽⼆级缓存是基于mapper⽂件的namespace的，也 就是说多个sqlSession可 以共享⼀个mapper中的⼆级缓存区域，并且如果两个mapper的namespace 相同，即使是两个mapper,那么这两 个mapper中执⾏sql查询到的数据也将存在相同的⼆级缓存区域 中

 如何使⽤⼆级缓存

（一）开启⼆级缓存

和⼀级缓存默认开启不⼀样，⼆级缓存需要我们⼿动开启

⾸先在全局配置⽂件sqlMapConfig.xml⽂件中加⼊如下代码:

<!--开启⼆级缓存-->
<settings>
 <setting name="cacheEnabled" value="true"/>
 </settings>

 其次在UserMapper.xml⽂件中开启缓存

<!--开启⼆级缓存-->
<cache></cache>

public class PerpetualCache implements Cache {
    private final String id;
    private MapcObject, Object> cache = new HashMapC);

    public PerpetualCache(St ring id) { this.id = id; }
}

 我们可以看到⼆级缓存底层还是HashMap结构

public class User implements Serializable(){
        //⽤户ID
        private int id;
        //⽤户姓名
        private String username;
        //⽤户性别
        private String sex; }

（三）测试

1. 测试⼆级缓存和sqlSession⽆关

@Test
    public void testTwoCache() {
        //根据 sqlSessionFactory 产⽣ session
        SqlSession sqlSession1 = sessionFactory.openSession();
        SqlSession sqlSession2 = sessionFactory.openSession();

        UserMapper userMapper1 = sqlSession1.getMapper(UserMapper.class);
        UserMapper userMapper2 = sqlSession2.getMapper(UserMapper.class);
        //第⼀次查询，发出sql语句，并将查询的结果放⼊缓存中
        User u1 = userMapper1.selectUserByUserId(1);
        System.out.println(u1);
        sqlSession1.close(); //第⼀次查询完后关闭 sqlSession

        //第⼆次查询，即使sqlSession1已经关闭了，这次查询依然不发出sql语句
        User u2 = userMapper2.selectUserByUserId(1);
        System.out.println(u2);
        sqlSession2.close();
    }

可以看出上⾯两个不同的sqlSession,第⼀个关闭了，第⼆次查询依然不发出sql查询语句

2. 测试执⾏commit()操作，⼆级缓存数据清空

@Test
    public void testTwoCache(){
        //根据 sqlSessionFactory 产⽣ session
        SqlSession sqlSession1 = sessionFactory.openSession();
        SqlSession sqlSession2 = sessionFactory.openSession();
        SqlSession sqlSession3 = sessionFactory.openSession();
        String statement = "com.lagou.pojo.UserMapper.selectUserByUserld" ;
        UserMapper userMapper1 = sqlSession1.getMapper(UserMapper. class );
        UserMapper userMapper2 = sqlSession2.getMapper(UserMapper. class );
        UserMapper userMapper3 = sqlSession2.getMapper(UserMapper. class );
        //第⼀次查询，发出sql语句，并将查询的结果放⼊缓存中
        User u1 = userMapperl.selectUserByUserId( 1 );
        System.out.println(u1);
        sqlSessionl .close(); //第⼀次查询完后关闭sqlSession

        //执⾏更新操作，commit()
        u1.setUsername( "aaa" );
        userMapper3.updateUserByUserId(u1);
        sqlSession3.commit();

        //第⼆次查询，由于上次更新操作，缓存数据已经清空(防⽌数据脏读)，这⾥必须再次发出sql语
        User u2 = userMapper2.selectUserByUserId( 1 );
        System.out.println(u2);
        sqlSession2.close();
    }

查看控制台情况：

4. useCache和flushCache

<select id="selectUserByUserId" useCache="false" resultType="com.lagou.pojo.User"
parameterType="int">
 select * from user where id=#{id}
</select>

<select id="selectUserByUserId" flushCache="true" useCache="false"
resultType="com.lagou.pojo.User" parameterType="int">
 select * from user where id=#{id}
</select>

三. ⼆级缓存整合redis

1. pom⽂件

<dependency>
 <groupId>org.mybatis.caches</groupId>
 <artifactId>mybatis-redis</artifactId>
 <version>1.0.0-beta2</version>
</dependency>

2. 配置⽂件

Mapper.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.lagou.mapper.IUserMapper"> <cache type="org.mybatis.caches.redis.RedisCache" />
<select id="findAll" resultType="com.lagou.pojo.User" useCache="true">
 select * from user
</select>

3. redis.properties

redis.host=localhost
redis.port=6379
redis.connectionTimeout=5000
redis.password=
redis.database=0

4. 测试

@Test
    public void SecondLevelCache(){
        SqlSession sqlSession1 = sqlSessionFactory.openSession();
        SqlSession sqlSession2 = sqlSessionFactory.openSession();
        SqlSession sqlSession3 = sqlSessionFactory.openSession();
        IUserMapper mapper1 = sqlSession1.getMapper(IUserMapper.class);
        lUserMapper mapper2 = sqlSession2.getMapper(lUserMapper.class);
        lUserMapper mapper3 = sqlSession3.getMapper(IUserMapper.class);
        User user1 = mapper1.findUserById(1);
        sqlSession1.close(); //清空⼀级缓存

        User user = new User();
        user.setId(1);
        user.setUsername("lisi");
        mapper3.updateUser(user);
        sqlSession3.commit();
        User user2 = mapper2.findUserById(1);
        System.out.println(user1==user2);
    }

源码分析：

public final class RedisCache implements Cache {
    public RedisCache(final String id) {
        if (id == null) {
            throw new IllegalArgumentException("Cache instances require anID");
        }
        this.id = id;
        RedisConfig redisConfig =
                RedisConfigurationBuilder.getInstance().parseConfiguration();
        pool = new JedisPool(redisConfig, redisConfig.getHost(),
                redisConfig.getPort(),
                redisConfig.getConnectionTimeout(),
                redisConfig.getSoTimeout(), redisConfig.getPassword(), redisConfig.getDatabase(),
                redisConfig.getClientName());
    }

public class RedisConfig extends JedisPoolConfig {
    private String host = Protocol.DEFAULT_HOST;
    private int port = Protocol.DEFAULT_PORT;
    private int connectionTimeout = Protocol.DEFAULT_TIMEOUT;
    private int soTimeout = Protocol.DEFAULT_TIMEOUT;
    private String password;
    private int database = Protocol.DEFAULT_DATABASE;
    private String clientName;
}

RedisConfig对象是由RedisConfigurationBuilder创建的，简单看下这个类的主要⽅法：

    public RedisConfig parseConfiguration(ClassLoader classLoader) {
        Properties config = new Properties();
        InputStream input =
                classLoader.getResourceAsStream(redisPropertiesFilename);
        if (input != null) {
            try {
                config.load(input);
            } catch (IOException e) {
                throw new RuntimeException(
                        "An error occurred while reading classpath property '"
                                + redisPropertiesFilename
                                + "', see nested exceptions", e);
            } finally {
                try {
                    input.close();
                } catch (IOException e) {
                    // close quietly
                }
            } 
        }
        RedisConfig jedisConfig = new RedisConfig();
        setConfigProperties(config, jedisConfig);
        return jedisConfig; 
}

核⼼的⽅法就是parseConfiguration⽅法，该⽅法从classpath中读取⼀个redis.properties⽂件:

host=localhost
port=6379
connectionTimeout=5000
soTimeout=5000
password= database=0 clientName=

    private Object execute(RedisCallback callback) {
        Jedis jedis = pool.getResource();
        try {
            return callback.doWithRedis(jedis);
        } finally {
            jedis.close();
        } 
}

模板接⼝为RedisCallback，这个接⼝中就只需要实现了⼀个doWithRedis⽅法⽽已：

public interface RedisCallback {
 Object doWithRedis(Jedis jedis);
}

    @Override
    public void putObject(final Object key, final Object value) {
        execute(new RedisCallback() {
            @Override
            public Object doWithRedis(Jedis jedis) {
                jedis.hset(id.toString().getBytes(), key.toString().getBytes(),
                        SerializeUtil.serialize(value));
                return null;
            }
        });
    }
    @Override
    public Object getObject(final Object key) {
        return execute(new RedisCallback() {

            @Override
            public Object doWithRedis(Jedis jedis) {
                return SerializeUtil.unserialize(jedis.hget(id.toString().getBytes(),
                        key.toString().getBytes()));
            }
        });
    }