SpringBoot配置多CacheManager

ゝ一世哀愁。 2022-02-01 00:25 1044阅读 0赞

## SpringCache配置多CacheManager ##

### 背景 ###

>  Spring为了减少数据的执行次数（重点在数据库查询方面）, 在其内部使用aspectJ技术，为执行操作的结果集做了一层缓存的抽象。这极大的提升了应用程序的性能。由于其切面注入的特性，所以不会对我们的程序造成任何的影响。对于一些实时性要求不那么高的业务数据，我们可以在Service上进行一些缓存的操作。这样就可以减少访问数据库的频率。

### 默认缓存的使用顺序 ###

>  在Spring内部，缓存的实现，依赖`org.springframework.cache.Cache`与`org.springframework.cache.CacheManager`共同协作，它们只是定义了一种规范接口，实际的存储规则，需要用户自己定义，当没有提供用户自定义Bean对象，SpringBoot会自动执行以下的检测顺序：

1.  [Generic][]
2.  [JCache (JSR-107)][JCache _JSR-107] (EhCache 3, Hazelcast, Infinispan, and others)
3.  [EhCache 2.x][]
4.  [Hazelcast][]
5.  [Infinispan][]
6.  [Couchbase][]
7.  [Redis][]
8.  [Caffeine][]
9.  [Simple][]
    
    **具体的用法请参考：[Spring Caching][]**

**Cache**

cache接口的作用是Spring提供的一种抽象操作规范，里面包含的`crud操作`

**CacheManager**

cacheManager接口的作用是用来获取Cache，类似一种对象工厂，所有的Cache，必须依赖与CacheManager来获取。

在Spring内部提供了三个默认的实现：

*  SimpleCacheManager 简单的集合实现
 *  ConcurrentMapCacheManager 内部使用ConcurrentHashMap作为缓存容器
 *  NoOpCacheManager 一个空的实现，不会保存任何记录。

当然其他的依赖需要我们引入三方的Jar，以及一些自定义的配置。

下面看下，执行切面的具体实现：

public class CacheInterceptor extends CacheAspectSupport implements MethodInterceptor, Serializable { 
    
       @Override
       @Nullable
       public Object invoke(final MethodInvocation invocation) throws Throwable { 
          Method method = invocation.getMethod();
    
          CacheOperationInvoker aopAllianceInvoker = () -> { 
             try { 
                return invocation.proceed();
             }
             catch (Throwable ex) { 
                throw new CacheOperationInvoker.ThrowableWrapper(ex);
             }
          };
    
          try { 
             return execute(aopAllianceInvoker, invocation.getThis(), method, invocation.getArguments());
          }
          catch (CacheOperationInvoker.ThrowableWrapper th) { 
             throw th.getOriginal();
          }
       }
    
    }

使用`AOP`的方式，将拦截添加缓存注解的方法，然后会调用父类的方法去执行：

`execute`方法中，会去寻找缓存的操作源`CacheOperationSource`，操作源中包含一个`CacheOperation`的集合。`CacheOperation`代表着`@Cachable`上注解信息的的实体类。然后我们可以用过`cacheManager`属性去寻找对应的`cacheManager`实现，获取`Cache`来完成缓存操作。

### 实现多CacheManager ###

前面分析了，缓存的大致实现过程，我们就可以使用`cacheManager`去有选择性的选择我们需要使用的缓存实现。

但是这里有个注意的点，我们需要给CacheManager一个默认的实现，这是由于Spring容器初始化机制造成的。`CacheAspectSupport`抽象实现了`SmartInitializingSingleton`接口，这个接口在Spring容器中，是单例对象的回调接口，当所有的单例对象实例化完毕，就会调用此方法。

我们观察下`CacheAspectSupport`的逻辑：

public void afterSingletonsInstantiated() { 
       if (getCacheResolver() == null) { 
          // Lazily initialize cache resolver via default cache manager...
          Assert.state(this.beanFactory != null, "CacheResolver or BeanFactory must be set on cache aspect");
          try { 
             setCacheManager(this.beanFactory.getBean(CacheManager.class));
          }
          catch (NoUniqueBeanDefinitionException ex) { 
             throw new IllegalStateException("No CacheResolver specified, and no unique bean of type " +
                   "CacheManager found. Mark one as primary or declare a specific CacheManager to use.");
          }
          catch (NoSuchBeanDefinitionException ex) { 
             throw new IllegalStateException("No CacheResolver specified, and no bean of type CacheManager found. " +
                   "Register a CacheManager bean or remove the @EnableCaching annotation from your configuration.");
          }
       }
       this.initialized = true;
    }

`setCacheManager(this.beanFactory.getBean(CacheManager.class));`，默认是从容器中拿到`CacheManager`对象，这就会出现一个问题，当配置多个`CacheManager`实例bean，并暴露给容器后，会出现Bean装载的错误，因为`getBean`，默认只会返回一个对象，当出现了两个Bean实例就会报错，找不到Bean对象。

所以，当对于自定义的多个Bean，我们需要指定一个打上`@Primary`注解，这样就可以解决冲突。

#### 具体实现 ####

接下来，贴出具体示例：

我们选择ehcache与redis作为多种缓存源。

**以下示例全部基于`2.1.4.RELEASE`版本，不同版本的代码差异较大**

1.  pom文件
    
        <dependency>
                    <groupId>net.sf.ehcache</groupId>
                    <artifactId>ehcache</artifactId>
                </dependency>
        
                <dependency>
                    <groupId>com.google.guava</groupId>
                    <artifactId>guava</artifactId>
                    <version>27.1-jre</version>
                </dependency>
          <dependency>
                    <groupId>org.springframework.boot</groupId>
                    <artifactId>spring-boot-starter-data-redis</artifactId>
                </dependency>
        <dependency>
                    <groupId>org.apache.commons</groupId>
                    <artifactId>commons-pool2</artifactId>
                </dependency>
          <dependency>
                    <groupId>com.h2database</groupId>
                    <artifactId>h2</artifactId>
                    <scope>runtime</scope>
                </dependency>
             <dependency>
                    <groupId>org.springframework.boot</groupId>
                    <artifactId>spring-boot-starter-data-jpa</artifactId>
                </dependency>
    
    我选择使用内置的H2+JPA作为数据持久层。因为这样比较方便。。不用连接MYSQL。
2.  application.yml配置
    
        spring:
          application:
            name: authority-management
          datasource:
            driver-class-name: org.h2.Driver
            url: jdbc:h2:~/test
            password:
            username: sa
          h2:
            console:
              enabled: true
          jpa:
            generate-ddl: true
            hibernate:
              ddl-auto: create-drop
            show-sql: true
          redis:
            host: XXXX
            port: 6379
            lettuce:
              pool:
                max-active: 8
                max-wait: 200
                max-idle: 8
                min-idle: 2
            timeout: 2000
          cache:
            ehcache:
              config: classpath:ehcache.xml
            type: EHCACHE
3.  编写配置类

@Configuration
    @EnableCaching
    @EnableConfigurationProperties(CacheProperties.class)
    public class CacheManagerConfiguration { 
        private final CacheProperties cacheProperties;
        public CacheManagerConfiguration(CacheProperties cacheProperties)    { 
            this.cacheProperties = cacheProperties;
        }
        public interface CacheManagerNames { 
            String REDIS_CACHE_MANAGER = "redisCacheManager";
            String EHCACHE_CACHE_MANAGER = "ehCacheManager";
        }
    
        @Bean(name = CacheManagerNames.REDIS_CACHE_MANAGER)
        public RedisCacheManager redisCacheManager(RedisConnectionFactory factory) { 
            Map<String, RedisCacheConfiguration> expires = ImmutableMap.<String, RedisCacheConfiguration>builder()
                    .put("15", RedisCacheConfiguration.defaultCacheConfig().entryTtl(
                            Duration.ofMillis(15)
                    ))
                    .put("30", RedisCacheConfiguration.defaultCacheConfig().entryTtl(
                            Duration.ofMillis(30)
                    ))
                    .put("60", RedisCacheConfiguration.defaultCacheConfig().entryTtl(
                            Duration.ofMillis(60)
                    ))
                    .put("120", RedisCacheConfiguration.defaultCacheConfig().entryTtl(
                            Duration.ofMillis(120)
                    ))
                    .build();
    
            RedisCacheManager redisCacheManager = RedisCacheManager.RedisCacheManagerBuilder.fromConnectionFactory(factory)
                    .withInitialCacheConfigurations(expires)
                    .build();
            return redisCacheManager;
        }
    
        @Bean(name = CacheManagerNames.EHCACHE_CACHE_MANAGER)
        @Primary
        public EhCacheCacheManager ehCacheManager() { 
            Resource resource = this.cacheProperties.getEhcache().getConfig();
            resource = this.cacheProperties.resolveConfigLocation(resource);
            EhCacheCacheManager ehCacheManager = new EhCacheCacheManager(
                    EhCacheManagerUtils.buildCacheManager(resource)
            );
            ehCacheManager.afterPropertiesSet();
            return ehCacheManager;
        }
    }

以上就是具体配置信息，需要主要的是别忘记`@Primary`

**Service实现**

@Override
        @Cacheable(key = "#userId", cacheNames = CacheManagerConfiguration.CacheNames.CACHE_15MINS,
                cacheManager = CacheManagerConfiguration.CacheManagerNames.EHCACHE_CACHE_MANAGER)
        public User findUserAccordingToId(Long userId) { 
            return userRepository.findById(userId).orElse(User.builder().build());
        }
    
        @Override
        @Cacheable(key = "#username", cacheNames = CacheManagerConfiguration.CacheNames.CACHE_15MINS,
                cacheManager = CacheManagerConfiguration.CacheManagerNames.REDIS_CACHE_MANAGER)
        public User findUserAccordingToUserName(String username) { 
            return userRepository.findUserByUsername(username);
        }

现在，就可以使用cacheManager属性来选择缓存源，用户可以灵活配置。

除了在方法上，使用注解外，我们还可以直接指定到类上，下面的示例，表示该类的全部方法都使用`Encache`作为缓存源。

@CacheConfig(cacheManager = CacheManagerConfiguration.CacheManagerNames.EHCACHE_CACHE_MANAGER)

1.  测试类+日志信息
    
        userService.findUserAccordingToId(save.getId());
        userService.findUserAccordingToId(save.getId());
        userService.findUserAccordingToUserName(save.getUsername());
        userService.findUserAccordingToUserName(save.getUsername());
    
        Hibernate: select user0_.id as id1_11_0_, user0_.email as email2_11_0_, user0_.image_url as image_ur3_11_0_, user0_.introduction as introduc4_11_0_, user0_.last_login_at as last_log5_11_0_, user0_.level as level6_11_0_, user0_.login_ip as login_ip7_11_0_, user0_.nickname as nickname8_11_0_, user0_.password as password9_11_0_, user0_.retry_login_count as retry_l10_11_0_, user0_.sex as sex11_11_0_, user0_.status as status12_11_0_, user0_.telephone as telepho13_11_0_, user0_.username as usernam14_11_0_ from user user0_ where user0_.id=?
        2019-05-12 20:43:06.486  INFO 12020 --- [           main] io.lettuce.core.EpollProvider            : Starting without optional epoll library
        2019-05-12 20:43:06.488  INFO 12020 --- [           main] io.lettuce.core.KqueueProvider           : Starting without optional kqueue library
        2019-05-12 20:43:06.807  INFO 12020 --- [           main] o.h.h.i.QueryTranslatorFactoryInitiator  : HHH000397: Using ASTQueryTranslatorFactory
        Hibernate: select user0_.id as id1_11_, user0_.email as email2_11_, user0_.image_url as image_ur3_11_, user0_.introduction as introduc4_11_, user0_.last_login_at as last_log5_11_, user0_.level as level6_11_, user0_.login_ip as login_ip7_11_, user0_.nickname as nickname8_11_, user0_.password as password9_11_, user0_.retry_login_count as retry_l10_11_, user0_.sex as sex11_11_, user0_.status as status12_11_, user0_.telephone as telepho13_11_, user0_.username as usernam14_11_ from user user0_ where user0_.username=?
        
        Process finished with exit code -1
    
     可以看到，两次查询，都只是用了一次select语句，还出现了一次`redis`连接操作，证明上述的配置是生效的。

#### 注意： ####

1.  我在此使用的`lettuce`作为redis客户端，使用连接池时，它依赖`commons-pool2`
2.  测试时，当redis作为缓存时，发现有时候还是会去数据库查询两次，怀疑是配置的超时时间问题。

[Generic]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-generic
[JCache _JSR-107]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-jcache
[EhCache 2.x]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-ehcache2
[Hazelcast]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-hazelcast
[Infinispan]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-infinispan
[Couchbase]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-couchbase
[Redis]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-redis
[Caffeine]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-caffeine
[Simple]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html#boot-features-caching-provider-simple
[Spring Caching]: https://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-caching.html