Redis Pipeline

Redis与客户端是通过TCP协议进行沟通的，这代表着执行一条命令，都和客户端建立tcp连接，执行命令，然后发返回结果，释放连接。如果在大量执行执行命令，其中的消耗也不可小觑。Redis引入了Pineline技术，实际上就是某一客户端把多个redis操作缓存起来，一并发送到redis执行，最后结果也打包返回。
Spring中的RedisTemplate也提供了对pipeline的支持，下面使用RedisTemplate做普通redis操作和pipeline操作的耗时对比：

package com.learning.systemservice;

import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.learning.entity.model.Community;
import com.learning.systemservice.dao.CommunityDao;
import lombok.extern.slf4j.Slf4j;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.data.redis.core.RedisCallback;
import org.springframework.data.redis.core.RedisTemplate;

import java.nio.charset.StandardCharsets;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;

@SpringBootTest
@Slf4j
class RedisPipeLineTest {
    @Autowired
    @Qualifier("jsonRedisTemplate")
    private RedisTemplate<String, String> redisTemplate;

    @Autowired
    private CommunityDao communityDao;

    @Autowired
    private ObjectMapper objectMapper;

    private Map<String, Community> collect;

    /**
     * 查询全国的社区信息，大约有六十多万，并转换成List<Entry<JsonString,Community>>的形式
     */
    @BeforeEach
    void queryData() {
        List<Community> communities = communityDao.queryAll(null);
        collect = communities.stream().collect(Collectors.toMap(record -> {
            try {
                return objectMapper.writeValueAsString(record);
            } catch (JsonProcessingException e) {
                e.printStackTrace();
            }
            return null;
        }, record -> record));
        log.info("total records:" + collect.size());
    }

    /**
     * 操作前清空redis数据库
     */
    @BeforeEach
    void clearRedisData() {
        redisTemplate.execute((RedisCallback<Object>) connection -> {
            connection.flushAll();
            return null;
        });
    }

    /**
     *  使用循环设置redis，key是community_id,value是json
     */
    @Test
    void TestRedisLoopSet() {
        long l1 = System.currentTimeMillis();
        collect.forEach((key, val) -> {
            redisTemplate.opsForValue().set("community_" + val.getId(), key);
        });
        long l2 = System.currentTimeMillis();
        log.info("time:" + (l2 - l1));
    }

    /**
     * 使用pipeline设置redis，key是community_id,value是json
     */
    @Test
    void TestRedisPipeLineSet() {
        long l1 = System.currentTimeMillis();
        redisTemplate.executePipelined((RedisCallback<Object>) connection -> {
            collect.forEach((json, obj) -> {
                connection.stringCommands().set(("community_" + obj.getId()).getBytes(StandardCharsets.UTF_8), json.getBytes(StandardCharsets.UTF_8));
            });
            return null;
        });
        long l2 = System.currentTimeMillis();
        log.info("time:" + (l2 - l1));
    }

}

循环set结果：488秒

...
2021-05-23 20:49:01.242  INFO 14384 --- [           main] com.zaxxer.hikari.HikariDataSource       : HikariPool-1 - Starting...
2021-05-23 20:49:02.585  INFO 14384 --- [           main] com.zaxxer.hikari.HikariDataSource       : HikariPool-1 - Start completed.
2021-05-23 20:49:06.329  INFO 14384 --- [           main] c.l.systemservice.RedisPipeLineTest      : total records:632599
2021-05-23 20:57:16.105  INFO 14384 --- [           main] c.l.systemservice.RedisPipeLineTest      : time:488267

pipeline结果：4.2秒

...
2021-05-23 21:04:40.439  INFO 9300 --- [           main] com.zaxxer.hikari.HikariDataSource       : HikariPool-1 - Starting...
2021-05-23 21:04:41.694  INFO 9300 --- [           main] com.zaxxer.hikari.HikariDataSource       : HikariPool-1 - Start completed.
2021-05-23 21:04:45.490  INFO 9300 --- [           main] c.l.systemservice.RedisPipeLineTest      : total records:632599
2021-05-23 21:04:51.264  INFO 9300 --- [           main] c.l.systemservice.RedisPipeLineTest      : time:4299