現(xiàn)開發(fā)的項(xiàng)目中决采，某接口需要頻獲取同一個(gè)實(shí)體，如果每次都查數(shù)據(jù)庫(kù)的話纵柿，性能消耗太大，于是決定使用redis做緩存启绰。
Redis默認(rèn)支持一般類型的數(shù)據(jù)昂儒，但需要對(duì)POJO進(jìn)行緩存的話，需要特殊處理委可，也就是需要對(duì)其進(jìn)行配置渊跋，使用jackson的ObjectMapper對(duì)實(shí)體進(jìn)行特殊處理。以下是編寫的第一版配置的代碼

import com.fasterxml.jackson.annotation.JsonAutoDetect
import com.fasterxml.jackson.annotation.PropertyAccessor
import com.fasterxml.jackson.databind.ObjectMapper
import com.fasterxml.jackson.databind.jsontype.impl.LaissezFaireSubTypeValidator
import org.springframework.beans.factory.annotation.Autowired
import org.springframework.boot.autoconfigure.data.redis.RedisProperties
import org.springframework.boot.context.properties.EnableConfigurationProperties
import org.springframework.cache.annotation.CachingConfigurerSupport
import org.springframework.cache.annotation.EnableCaching
import org.springframework.context.annotation.Bean
import org.springframework.context.annotation.Configuration
import org.springframework.data.redis.connection.RedisConnectionFactory
import org.springframework.data.redis.connection.RedisStandaloneConfiguration
import org.springframework.data.redis.connection.lettuce.LettuceConnectionFactory
import org.springframework.data.redis.core.RedisTemplate
import org.springframework.data.redis.serializer.Jackson2JsonRedisSerializer
import org.springframework.data.redis.serializer.StringRedisSerializer

/**
 * @author Anson
 * @date 2020/11/7
 */
@EnableCaching
@Configuration
@EnableConfigurationProperties(
        RedisProperties::class
)
class RedisConfig @Autowired constructor(
        private val properties: RedisProperties
): CachingConfigurerSupport() {

    @Bean
    fun redisConnectionFactory(): LettuceConnectionFactory {
        val config = RedisStandaloneConfiguration()
        config.hostName = properties.host
        config.port = properties.port
        return LettuceConnectionFactory(config)
    }
    @Bean
    fun cacheManager(template: RedisTemplate<String, Any>): CacheManager {
        val defaultCacheConfiguration = RedisCacheConfiguration
                .defaultCacheConfig() // 設(shè)置key為String
                .serializeKeysWith(RedisSerializationContext.SerializationPair.fromSerializer(template.stringSerializer)) // 設(shè)置value 為自動(dòng)轉(zhuǎn)Json的Object
                .serializeValuesWith(RedisSerializationContext.SerializationPair.fromSerializer(template.valueSerializer)) // 不緩存null
                .disableCachingNullValues() // 緩存數(shù)據(jù)保存2小時(shí)
                .entryTtl(Duration.ofHours(2))
        return RedisCacheManager.RedisCacheManagerBuilder // Redis 連接工廠
                .fromConnectionFactory(template.connectionFactory!!) // 緩存配置
                .cacheDefaults(defaultCacheConfiguration) // 配置同步修改或刪除 put/evict
                .transactionAware()
                .build()
    }

    private fun jackson2JsonRedisSerializer(): Jackson2JsonRedisSerializer<Any> {
        val jackson2JsonRedisSerializer = Jackson2JsonRedisSerializer(Any::class.java)
        val objectMapper = ObjectMapper()
        objectMapper.setVisibility(PropertyAccessor.ALL, JsonAutoDetect.Visibility.ANY)
        jackson2JsonRedisSerializer.setObjectMapper(objectMapper)
        return jackson2JsonRedisSerializer
    }

    @Bean
    fun redisTemplate(factory: RedisConnectionFactory): RedisTemplate<String, Any> {
        val template = RedisTemplate<String, Any>()
        // 配置連接工廠
        template.setConnectionFactory(factory)
        // 定義Jackson2JsonRedisSerializer序列化對(duì)象
        val jacksonSeial = jackson2JsonRedisSerializer()
        val stringSerial = StringRedisSerializer()
        // redis key 序列化方式使用stringSerial
        template.keySerializer = stringSerial
        // redis value 序列化方式使用jackson
        template.valueSerializer = jacksonSeial
        // redis hash key 序列化方式使用stringSerial
        template.hashKeySerializer = stringSerial
//        // redis hash value 序列化方式使用jackson
        template.hashValueSerializer = jacksonSeial
        template.afterPropertiesSet()
        return template
    }
}

簡(jiǎn)單的團(tuán)隊(duì)POJO着倾，用于從數(shù)據(jù)庫(kù)中查出來拾酝，并放在緩存

data class Team(
        var id: String? = null,
        val name: String? = null,
        var enabled: Boolean? = true,
        var rf: Boolean? = false
)

團(tuán)隊(duì)Service，緩存分區(qū)配置屈呕，獲取團(tuán)隊(duì)與更新團(tuán)隊(duì)

@Service
@Transactional
@CacheConfig(cacheNames = ["teams"])
class TeamServiceImpl(private val mapper: TeamMapper) {

    @Cacheable(key = "#id", unless = "#result==null")
    suspend fun getById(id: String): Team? {
        return mapper.getById(id)
    }

    @CacheEvict(key = "#id")
    suspend fun update(id: String, t: Team): Team {
        return mapper.update(t).let { t }
    }
}

坑一

問題

通過getById獲取team數(shù)據(jù)后微宝，會(huì)通過注解@Cacheable把數(shù)據(jù)以Key-Value形式放入緩存，但通過日志發(fā)現(xiàn)Spring自動(dòng)生成的key貌似和預(yù)期不符虎眨，我們理想中的key格式應(yīng)該是teams::04422030192c*******8af8c684740bb蟋软。所以使用@CacheEvict注解怎么也無法刪除之前的緩存镶摘。

$3
SET
$153
teams::[04422030192c*******8af8c684740bb,Continuation at com.***.controllers.TeamController$findById$1.invokeSuspend(TeamController.kt:27)]

原因

Spring默認(rèn)使用的是SimpleKeyGenerator去生成數(shù)據(jù)的key，從以下源碼可以看出岳守，它會(huì)把所有參數(shù)params丟到SimpleKey中進(jìn)行組合凄敢，生成上日志中的形式。

public class SimpleKeyGenerator implements KeyGenerator {
    public SimpleKeyGenerator() {
    }

    public Object generate(Object target, Method method, Object... params) {
        return generateKey(params);
    }

    public static Object generateKey(Object... params) {
        if (params.length == 0) {
            return SimpleKey.EMPTY;
        } else {
            if (params.length == 1) {
                Object param = params[0];
                if (param != null && !param.getClass().isArray()) {
                    return param;
                }
            }

            return new SimpleKey(params);
        }
    }
}

解決

因此不能使用Spring自帶的SimpleKeyGenerator湿痢，只能自己寫一個(gè)涝缝。

1. 在RedisConfig加入自定義的 idKeyGenerator。

@Bean
fun idKeyGenerator(): KeyGenerator {
     return KeyGenerator { _, _, params ->
         val sb = StringBuilder().append(params[0])
         sb.toString()
     }
}

2. 只需修改@Cacheable內(nèi)參數(shù)譬重，使用自定義的idKeyGenerator

@Cacheable(keyGenerator = "idKeyGenerator", unless = "#result==null")
suspend fun getById(id: String): Team? {
    return mapper.getById(id)
}

成功插入拒逮、查詢與刪除緩存

$3
SET
$42
teams::04422030192c*******8af8c684740bb
// 省略
$3
GET
$42
teams::04422030192c*******8af8c684740bb
// 省略
$3
DEL
$42
teams::04422030192c*******8af8c684740bb

坑二

問題

需對(duì)LocalDateTime類型支持

解決

修改配置，加入以下代碼

val javaTimeModule = JavaTimeModule()
javaTimeModule.addSerializer(LocalDateTime::class.java, LocalDateTimeSerializer(DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm:ss")))
objectMapper.registerModule(javaTimeModule)

坑三

問題

第一次獲取數(shù)據(jù)都是直接從數(shù)據(jù)庫(kù)查的臀规，所以都沒有問題滩援，但第二次查詢則是從緩存里面獲取。于是拋以下異常塔嬉。

java.lang.ClassCastException: java.util.LinkedHashMap cannot be cast to com.XXX

找了好久的原因還有debug玩徊，發(fā)現(xiàn)是在之前的配置Jackson2JsonRedisSerializer里的參數(shù)是Any::class.java，因此把數(shù)據(jù)轉(zhuǎn)成LinkedHashMap谨究，因此無論如何也無法轉(zhuǎn)成我們需要的實(shí)體恩袱。在網(wǎng)上也搜了好久，說給ObjectMapper加一行配置objectMapper.enableDefaultTyping(ObjectMapper.DefaultTyping.NON_FINAL)即可胶哲，但enableDefaultTyping該方法已經(jīng)過時(shí)畔塔，不再建議使用，正確的方法是下行代碼纪吮。

objectMapper.activateDefaultTyping(LaissezFaireSubTypeValidator.instance, ObjectMapper.DefaultTyping.NON_FINAL)

網(wǎng)上有介紹該行代碼的介紹俩檬，附上網(wǎng)址。但是加上后依然報(bào)錯(cuò)碾盟。

org.springframework.data.redis.serializer.SerializationException: Could not read JSON: Unexpected token (START_OBJECT), expected START_ARRAY: need JSON Array to contain As.WRAPPER_ARRAY type information for class java.lang.Object
 at [Source: (byte[])"{"id":"0510c38023********c7c7a8747f6e40","[truncated 32 bytes]; line: 1, column: 1]; nested exception is com.fasterxml.jackson.databind.exc.MismatchedInputException: Unexpected token (START_OBJECT), expected START_ARRAY: need JSON Array to contain As.WRAPPER_ARRAY type information for class java.lang.Object

原因

加上那行代碼棚辽，只會(huì)對(duì)除了一些自然類型(String、Double冰肴、Integer屈藐、Double)類型外的非常量(non-final)類型加上值類型，但我們的實(shí)體類他并沒有加上類型熙尉，因此他仍然以Json形式繼續(xù)轉(zhuǎn)換联逻，但由于存到緩存里的數(shù)據(jù)已不是正常的Json了，因此導(dǎo)致轉(zhuǎn)換錯(cuò)誤检痰。

解決

只需把ObjectMapper.DefaultTyping.NON_FINAL)改成ObjectMapper.DefaultTyping.EVERYTHING)就解決了我們的問題包归。

最后的jackson代碼塊

private fun jackson2JsonRedisSerializer(): Jackson2JsonRedisSerializer<Any> {
        val jackson2JsonRedisSerializer = Jackson2JsonRedisSerializer(Any::class.java)
        val objectMapper = ObjectMapper()
        objectMapper.setVisibility(PropertyAccessor.ALL, JsonAutoDetect.Visibility.ANY)
        objectMapper.activateDefaultTyping(LaissezFaireSubTypeValidator.instance, ObjectMapper.DefaultTyping.EVERYTHING)
        val javaTimeModule = JavaTimeModule()
        javaTimeModule.addSerializer(LocalDateTime::class.java, LocalDateTimeSerializer(DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm:ss")))
        objectMapper.registerModule(javaTimeModule)
        jackson2JsonRedisSerializer.setObjectMapper(objectMapper)
        return jackson2JsonRedisSerializer
}