一、介紹

當(dāng)一種請求，總是能越過緩存，調(diào)用數(shù)據(jù)庫，就是緩存穿透。

比如當(dāng)請求一個數(shù)據(jù)庫沒有的數(shù)據(jù)，那么緩存也不會有，然后就一直請求，甚至高并發(fā)去請求，對數(shù)據(jù)庫壓力會增大。

二、方案介紹

1. 如果key具有某種規(guī)則，那么可以對key增加校驗(yàn)機(jī)制，不符合直接返回。
2. Redisson布隆過濾器
3. 邏輯修改，當(dāng)數(shù)據(jù)庫沒有此數(shù)據(jù)，以null為value，也插入redis緩存，但設(shè)置較短的過期時間。

三、Redis Docker部署

docker-compose示例如下，redis.conf從這里下載

redis:
    container_name: redis
    image: redis:7.2
    volumes:
      - ./redis/redis.conf:/usr/local/etc/redis/redis.conf
    ports:
      - "6379:6379"
    command: [ "redis-server", "/usr/local/etc/redis/redis.conf" ]

四、SpringBoot3 Base代碼

1. 依賴配置

		<!-- redis -->
		<dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-data-redis</artifactId>
        </dependency>
        <!-- redis 連接線程池 -->
        <dependency>
            <groupId>org.apache.commons</groupId>
            <artifactId>commons-pool2</artifactId>
            <version>2.11.1</version>
        </dependency>
        <!-- redisson -->
        <dependency>
            <groupId>org.redisson</groupId>
            <artifactId>redisson</artifactId>
            <version>3.24.3</version>
        </dependency>

spring: 
  data:
    redis:
      host: 192.168.101.65  # Redis服務(wù)器的主機(jī)名或IP地址
      port: 6379  # Redis服務(wù)器的端口號
      password:  # 用于連接Redis服務(wù)器的密碼
      database: 0  # 要連接的Redis數(shù)據(jù)庫的索引號
      lettuce:
        pool:
          max-active: 20  # 連接池中最大的活躍連接數(shù)
          max-idle: 10  # 連接池中最大的空閑連接數(shù)
          min-idle: 0  # 連接池中最小的空閑連接數(shù)
      timeout: 10000  # 連接超時時間（毫秒）
      lock-watchdog-timeout: 100  # Redisson的分布式鎖的看門狗超時時間（毫秒）

2. 基本代碼

要演示的代碼很簡單，就是一個攜帶courseId請求過來，調(diào)用下面的service函數(shù)，然后查詢數(shù)據(jù)庫。

@Override
    public CoursePublish getCoursePublish(Long courseId) {
        return coursePublishMapper.selectById(courseId);
    }

當(dāng)我們使用redis改造時，基本代碼如下

@Override
    public CoursePublish getCoursePublishCache(Long courseId) {
        String key = "content:course:publish:" + courseId;
        //先查詢redis
        Object object = redisTemplate.opsForValue().get(key);
        if (object != null){
            String string = object.toString();
            CoursePublish coursePublish = JSON.parseObject(string, CoursePublish.class);
            return coursePublish;
        }else {
            //后查詢數(shù)據(jù)庫
            CoursePublish coursePublish = getCoursePublish(courseId);
            if (coursePublish != null){
                redisTemplate.opsForValue().set(key, JSON.toJSONString(coursePublish));
            }
            return coursePublish;
        }
    }

五、緩存優(yōu)化代碼

1. 校驗(yàn)機(jī)制

我這里的id沒規(guī)則，所以加不了，跳過。

2. 布隆過濾器

讀取yaml配置

@Data
@Component
@ConfigurationProperties(prefix = "spring.data.redis")
public class RedisProperties {
    private String host;
    private int port;
    private String password;
    private int database;
    private int lockWatchdogTimeout;
}

配置RedissonClient

@Slf4j
@Configuration
public class RedissionConfig {

    @Autowired
    private RedisProperties redisProperties;

    @Bean
    public RedissonClient redissonClient() {
        RedissonClient redissonClient;

        Config config = new Config();
        //starter依賴進(jìn)來的redisson要以redis://開頭，其他不用
        String url = "redis://"+ redisProperties.getHost() + ":" + redisProperties.getPort();
        config.useSingleServer().setAddress(url)
                //.setPassword(redisProperties.getPassword())
                .setDatabase(redisProperties.getDatabase());

        try {
            redissonClient = Redisson.create(config);
            return redissonClient;
        } catch (Exception e) {
            log.error("RedissonClient init redis url:[{}], Exception:", url, e);
            return null;
        }
    }
}

把布隆過濾器加到service，如下

	private RBloomFilter<String> bloomFilter;

    @PostConstruct
    public void init(){
        //初始化布隆過濾器
        bloomFilter = redissonClient.getBloomFilter("bloom-filter");
        bloomFilter.tryInit(100, 0.003);
        List<CoursePublish> coursePublishList = coursePublishMapper.selectList(new LambdaQueryWrapper<CoursePublish>());
        coursePublishList.forEach(coursePublish -> {
            String key = "content:course:publish:" + coursePublish.getId();
            bloomFilter.add(key);
        });
    }
    
	@Override
    public CoursePublish getCoursePublishCache(Long courseId) {
        String key = "content:course:publish:" + courseId;
        //布隆過濾器
        boolean contains = bloomFilter.contains(key);
        if (!contains){
            return null;
        }
        //先查詢redis
        Object object = redisTemplate.opsForValue().get(key);
        if (object != null){
            String string = object.toString();
            CoursePublish coursePublish = JSON.parseObject(string, CoursePublish.class);
            return coursePublish;
        }else {
            //后查詢數(shù)據(jù)庫
            CoursePublish coursePublish = getCoursePublish(courseId);
            if (coursePublish != null){
            	bloomFilter.add(key);
                redisTemplate.opsForValue().set(key, JSON.toJSONString(coursePublish));
            }
            return coursePublish;
        }
    }

3. 邏輯優(yōu)化

當(dāng)數(shù)據(jù)庫沒有此數(shù)據(jù)，以null為value，也插入redis緩存，但設(shè)置較短的過期時間。

            //后查詢數(shù)據(jù)庫
            CoursePublish coursePublish = getCoursePublish(courseId);
            if (coursePublish != null) {
            	bloomFilter.add(key);
                redisTemplate.opsForValue().set(key, JSON.toJSONString(coursePublish));
            }else {
                redisTemplate.opsForValue().set(key, JSON.toJSONString(coursePublish), 10, TimeUnit.SECONDS);
            }
            return coursePublish;

文章來源地址http://www.zghlxwxcb.cn/news/detail-834269.html

到了這里，關(guān)于Redis之緩存穿透問題解決方案實(shí)踐SpringBoot3+Docker的文章就介紹完了。如果您還想了解更多內(nèi)容，請?jiān)谟疑辖撬阉鱐OY模板網(wǎng)以前的文章或繼續(xù)瀏覽下面的相關(guān)文章，希望大家以后多多支持TOY模板網(wǎng)！