JetCache二级缓存使用

一起养成写作习惯！这是我参与「掘金日新计划 · 4 月更文挑战」的第5天，点击查看活动详情。

简介

JetCache是一个基于Java的缓存系统封装，提供统一的API和注解来简化缓存的使用。 JetCache提供了比SpringCache更加强大的注解，可以原生的支持TTL、两级缓存、分布式自动刷新，还提供了Cache接口用于手工缓存操作。 当前有四个实现，RedisCache、TairCache（此部分未在github开源）、CaffeineCache(in memory)和一个简易的LinkedHashMapCache(in memory)，要添加新的实现也是非常简单的。

全部特性:

• 通过统一的API访问Cache系统
• 通过注解实现声明式的方法缓存，支持TTL和两级缓存
• 通过注解创建并配置Cache实例
• 针对所有Cache实例和方法缓存的自动统计
• Key的生成策略和Value的序列化策略是可以配置的
• 分布式缓存自动刷新，分布式锁 (2.2+)
• 异步Cache API (2.2+，使用Redis的lettuce客户端时)
• Spring Boot支持

基本配置（使用Spring Boot）

添加依赖

<dependency>
    <groupId>com.alicp.jetcache</groupId>
    <artifactId>jetcache-starter-redis-lettuce</artifactId>
    <version>2.6.3</version>
</dependency>
复制代码

配置一个spring boot风格的application.yml文件，把他放到资源目录中

jetcache:
  statIntervalMinutes: 60
  areaInCacheName: false
  hidePackages: com.f
  local:
    default:
      type: caffeine
      keyConvertor: fastjson
      limit: 256
      expireAfterWriteInMillis: 500
  remote:
    default:
      type: redis.lettuce
      expireAfterWriteInMillis: 5000
      keyConvertor: fastjson
      valueEncoder: kryo
      valueDecoder: kryo
      uri: redis://192.168.137.1:6379/1?timeout=100ms
复制代码

添加配置类EnableMethodCache，EnableCreateCacheAnnotation这两个注解分别激活Cached和CreateCache注解。

@EnableMethodCache(basePackages = "com.f")
@EnableCreateCacheAnnotation
@Configuration
public class CacheConfig {

    @Bean(name = "redisClient")
    @DependsOn(RedisLettuceAutoConfiguration.AUTO_INIT_BEAN_NAME)
    public LettuceFactory redisClient() {
        return new LettuceFactory("remote.default", RedisClient.class);
    }

}
复制代码

@CreateCache使用，以及获取原生redis连接

@Component
@Accessors(fluent = true)
@Getter
@Slf4j
public class CacheTemplate {

    private final RedisClient redisClient;

    private final StatefulRedisConnection<String, String> connect;

    private final StatefulRedisConnection<String, Object> objectConnection;

    private final RedisCommands<String, String> sync;

    private final RedisAsyncCommands<String, String> async;

    private final RedisReactiveCommands<String, String> reactive;

    private final RedisCommands<String, Object> syncObject;

    private final RedisAsyncCommands<String, Object> asyncObject;

    private final RedisReactiveCommands<String, Object> reactiveObject;

    public CacheTemplate(final RedisClient redisClient) {
        this.redisClient = redisClient;
        this.connect = redisClient.connect();
        this.objectConnection = redisClient.connect(KryoCodec.INSTANCE);
        this.sync = connect.sync();
        this.async = connect.async();
        this.reactive = connect.reactive();
        this.syncObject = objectConnection.sync();
        this.asyncObject = objectConnection.async();
        this.reactiveObject = objectConnection.reactive();
    }

    @PostConstruct
    public void pingRedis() {
        new ScheduledThreadPoolExecutor(1, new NamedThreadFactory("redis-ping")).scheduleAtFixedRate(() -> log.debug("ping:{}", sync.ping()), 15, 150, TimeUnit.SECONDS);
    }

    @CreateCache(name = "common", expire = 30, cacheType = CacheType.REMOTE)
    private Cache<String, Object> commonCache;

    @CreateCache(name = "local", expire = 3, localLimit = 8096, cacheType = CacheType.LOCAL)
    private Cache<String, Object> localCache;

    @PreDestroy
    public void destroy() {
        connect.close();
        objectConnection.close();
        redisClient.shutdown();
    }
}
复制代码

@Cached注解使用

@Cached(name = "perms", key = "#userId", localLimit = 1024, localExpire = 5, expire = 120, cacheType = CacheType.BOTH)
@Override
public Set<String> selectPermsByUserId(Long userId) {
    return getBaseMapper().selectPermsByUserId(userId);
}
复制代码

参考

• wiki
• f-cloud