SpringBoot整合redis客戶端超時(shí)怎么解決

本文小編為大家詳細(xì)介紹“SpringBoot整合redis客戶端超時(shí)怎么解決”，內(nèi)容詳細(xì)，步驟清晰，細(xì)節(jié)處理妥當(dāng)，希望這篇“SpringBoot整合redis客戶端超時(shí)怎么解決”文章能幫助大家解決疑惑，下面跟著小編的思路慢慢深入，一起來學(xué)習(xí)新知識吧。

問題

在開發(fā)的時(shí)候，使用到Lettuce連接redis，一段時(shí)間后不操作，再去操作redis，會報(bào)連接超時(shí)錯(cuò)誤，在其重連后又可使用。

原因是：Lettuce 自適應(yīng)拓?fù)渌⑿拢ˋdaptive updates）與定時(shí)拓?fù)渌⑿拢≒eriodic updates） 是默認(rèn)關(guān)閉的導(dǎo)致問題的出現(xiàn)

解決的方案

方法一：

1、重寫連接工廠實(shí)例，更改其LettuceClientConfiguration 為開啟拓?fù)涓?/p>

@Configuration
public class RedisConfig {
 
 
    @Autowired
    private RedisProperties redisProperties;
 
    //這是固定的模板
    //自己定義了一個(gè)RedisTemplate
    @Bean
    @SuppressWarnings("all")
    public RedisTemplate<String, Object> redisTemplate(@Qualifier("lettuceConnectionFactoryUvPv") RedisConnectionFactory factory) {
        RedisTemplate<String, Object> template = new RedisTemplate<>();
        template.setConnectionFactory(factory);
 
        //Json序列化配置
        Jackson2JsonRedisSerializer<Object> jackson2JsonRedisSerializer = new Jackson2JsonRedisSerializer<Object>(Object.class);
        ObjectMapper om = new ObjectMapper();
        om.setVisibility(PropertyAccessor.ALL, JsonAutoDetect.Visibility.ANY);
        om.activateDefaultTyping(om.getPolymorphicTypeValidator());
        om.enableDefaultTyping(ObjectMapper.DefaultTyping.NON_FINAL);
        //解決序列化問題
        om.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
        jackson2JsonRedisSerializer.setObjectMapper(om);
 
        //String的序列化
        StringRedisSerializer stringRedisSerializer = new StringRedisSerializer();
 
        //key采用String的序列化方式
        template.setKeySerializer(stringRedisSerializer);
        //hash的key也采用String的序列化方式
        template.setHashKeySerializer(stringRedisSerializer);
 
        //value序列化方式采用jackson
        template.setValueSerializer(jackson2JsonRedisSerializer);
 
        //hash的value序列化方式采用jackson
        template.setHashValueSerializer(jackson2JsonRedisSerializer);
        template.afterPropertiesSet();
 
        return template;
    }
 
    /**
     * 為RedisTemplate配置Redis連接工廠實(shí)現(xiàn)
     * LettuceConnectionFactory實(shí)現(xiàn)了RedisConnectionFactory接口
     * UVPV用Redis
     *
     * @return 返回LettuceConnectionFactory
     */
    @Bean(destroyMethod = "destroy")
    //這里要注意的是，在構(gòu)建LettuceConnectionFactory 時(shí)，如果不使用內(nèi)置的destroyMethod，可能會導(dǎo)致Redis連接早于其它Bean被銷毀
    public LettuceConnectionFactory lettuceConnectionFactoryUvPv() throws Exception {
 
//        List<String> clusterNodes = redisProperties.getCluster().getNodes();
//        Set<RedisNode> nodes = new HashSet<>();
//        clusterNodes.forEach(address -> nodes.add(new RedisNode(address.split(":")[0].trim(), Integer.parseInt(address.split(":")[1]))));
//        RedisClusterConfiguration clusterConfiguration = new RedisClusterConfiguration();
//        clusterConfiguration.setClusterNodes(nodes);
//        clusterConfiguration.setPassword(RedisPassword.of(redisProperties.getPassword()));
//        clusterConfiguration.setMaxRedirects(redisProperties.getCluster().getMaxRedirects());
 
        //我使用的是單機(jī)redis，集群使用上面注釋的代碼
        Set<RedisNode> nodes = new HashSet<>();
        nodes.add(new RedisNode(redisProperties.getHost(), redisProperties.getPort()));
 
 
        RedisStandaloneConfiguration  redisStandaloneConfiguration=new RedisStandaloneConfiguration();
        redisStandaloneConfiguration.setHostName(redisProperties.getHost());
        redisStandaloneConfiguration.setPassword(redisProperties.getPassword());
        redisStandaloneConfiguration.setDatabase(redisProperties.getDatabase());
        redisStandaloneConfiguration.setPort(redisProperties.getPort());
 
        GenericObjectPoolConfig poolConfig = new GenericObjectPoolConfig();
        poolConfig.setMaxIdle(redisProperties.getLettuce().getPool().getMaxIdle());
        poolConfig.setMinIdle(redisProperties.getLettuce().getPool().getMinIdle());
        poolConfig.setMaxTotal(redisProperties.getLettuce().getPool().getMaxActive());
 
        return new LettuceConnectionFactory(redisStandaloneConfiguration, getLettuceClientConfiguration(poolConfig));
    }
 
    /**
     * 配置LettuceClientConfiguration 包括線程池配置和安全項(xiàng)配置
     *
     * @param genericObjectPoolConfig common-pool2線程池
     * @return lettuceClientConfiguration
     */
    private LettuceClientConfiguration getLettuceClientConfiguration(GenericObjectPoolConfig genericObjectPoolConfig) {
        /*
        ClusterTopologyRefreshOptions配置用于開啟自適應(yīng)刷新和定時(shí)刷新。如自適應(yīng)刷新不開啟，Redis集群變更時(shí)將會導(dǎo)致連接異常！
         */
        ClusterTopologyRefreshOptions topologyRefreshOptions = ClusterTopologyRefreshOptions.builder()
                //開啟自適應(yīng)刷新
                //.enableAdaptiveRefreshTrigger(ClusterTopologyRefreshOptions.RefreshTrigger.MOVED_REDIRECT, ClusterTopologyRefreshOptions.RefreshTrigger.PERSISTENT_RECONNECTS)
                //開啟所有自適應(yīng)刷新，MOVED，ASK，PERSISTENT都會觸發(fā)
                .enableAllAdaptiveRefreshTriggers()
                // 自適應(yīng)刷新超時(shí)時(shí)間(默認(rèn)30秒)
                .adaptiveRefreshTriggersTimeout(Duration.ofSeconds(25)) //默認(rèn)關(guān)閉開啟后時(shí)間為30秒
                // 開周期刷新
                .enablePeriodicRefresh(Duration.ofSeconds(20))  // 默認(rèn)關(guān)閉開啟后時(shí)間為60秒 ClusterTopologyRefreshOptions.DEFAULT_REFRESH_PERIOD 60  .enablePeriodicRefresh(Duration.ofSeconds(2)) = .enablePeriodicRefresh().refreshPeriod(Duration.ofSeconds(2))
                .build();
        return LettucePoolingClientConfiguration.builder()
                .poolConfig(genericObjectPoolConfig)
                .clientOptions(ClusterClientOptions.builder().topologyRefreshOptions(topologyRefreshOptions).build())
                //將appID傳入連接，方便Redis監(jiān)控中查看
                //.clientName(appName + "_lettuce")
                .build();
    }
 
}

2、SpringBoot2.3.x后，可使用配置文件中開啟lettuce的拓?fù)渌⑿?/p>

lettuce:
    pool:
      max-active: 20
      max-wait: -1ms
      max-idle: 10
      min-idle: 2
    cluster:
      refresh:
        adaptive: true
        #20秒自動(dòng)刷新一次
        period: 20

方法二：

更改連接redis的連接方式，使用jedis連接

<dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-data-redis</artifactId>
            <exclusions>
                <exclusion>
                    <groupId>io.lettuce</groupId>
                    <artifactId>lettuce-core</artifactId>
                </exclusion>
            </exclusions>
        </dependency>
         
        <dependency>
            <groupId>redis.clients</groupId>
            <artifactId>jedis</artifactId>
        </dependency>

配置文件

spring:
  redis:
    password: xxx
    host: 172.16.0.x
    port: 6579
    timeout: 5000
    jedis:
      pool:
              #最大連接數(shù)據(jù)庫連接數(shù),設(shè) 0 為沒有限制
        max-active: 8
              #最大等待連接中的數(shù)量,設(shè) 0 為沒有限制
        max-idle: 8
              #最大建立連接等待時(shí)間。如果超過此時(shí)間將接到異常。設(shè)為-1表示無限制。
        max-wait: -1ms
              #最小等待連接中的數(shù)量,設(shè) 0 為沒有限制
        min-idle: 0
    #lettuce:
      #pool:
        #max-active: ${redis.config.maxTotal:1024}
        #max-idle: ${redis.config.maxIdle:50}
        #min-idle: ${redis.config.minIdle:1}
        #max-wait: ${redis.config.maxWaitMillis:5000}

讀到這里，這篇“SpringBoot整合redis客戶端超時(shí)怎么解決”文章已經(jīng)介紹完畢，想要掌握這篇文章的知識點(diǎn)還需要大家自己動(dòng)手實(shí)踐使用過才能領(lǐng)會，如果想了解更多相關(guān)內(nèi)容的文章，歡迎關(guān)注億速云行業(yè)資訊頻道。