✨✨个人主页:沫洺的主页

📚📚系列专栏: 📖 JavaWeb专栏📖 JavaSE专栏 📖 Java基础专栏📖vue3专栏 

                           📖MyBatis专栏📖Spring专栏📖SpringMVC专栏📖SpringBoot专栏

                           📖Docker专栏📖Reids专栏📖MQ专栏📖SpringCloud专栏     

💖💖如果文章对你有所帮助请留下三连✨✨

💌RedisTemplate使用Scan 

@SpringBootTest
public class AppTests_Scan {@Resource(name = "redisTemplate")private ValueOperations valueOperations;@Autowiredprivate StringRedisTemplate stringRedisTemplate;@Testvoid test1() {for (int i =1 ;i<11;i++){User user = User.builder().name(i+"号张").age(18).build();valueOperations.set("user."+i,user);}}}

原生的scan用法

RedisTemplate使用Scan 

@SpringBootTest
public class AppTests_Scan {@Resource(name = "redisTemplate")private ValueOperations valueOperations;@Autowiredprivate StringRedisTemplate stringRedisTemplate;@Testvoid test1() {for (int i =1 ;i<11;i++){User user = User.builder().name(i+"号张").age(18).build();valueOperations.set("user."+i,user);}}@Testvoid test2(){//迭代扫描,指定模糊匹配,预想数量ScanOptions scanOptions = ScanOptions.scanOptions().match("user.*").count(3).build();//获取游标Cursor cursor = stringRedisTemplate.scan(scanOptions);while (cursor.hasNext()){//获取游标idSystem.out.println(cursor.getCursorId());System.out.println(cursor.next());}cursor.close();}
}

💌模拟多线程并发使用Decrby 

模拟高并发场景线上购物,多个用户(线程)同时选购某产品,提示库存信息

先初始化产品库存

@SpringBootTest
class AppTests_DecrBy {@Resource(name = "redisTemplate")private ValueOperations valueOperations;@Autowiredprivate StringRedisTemplate stringRedisTemplate;//定义一个产品private final static String productKey = "product01";@Testvoid test1() {//设置初始化库存5个valueOperations.set(productKey, 5);}
}

不考虑并发的情况下

    @Testvoid test2() throws InterruptedException {//获取线程池ExecutorService executorService = Executors.newCachedThreadPool();//模拟开启10个线程for (int i = 1; i <= 10; i++) {//开启线程执行executorService.execute(() -> {//获取库存Integer qty = valueOperations.get(productKey);//如果库存充足,产品数量就减一if (qty > 0) {qty = qty - 1;//重新设置库存valueOperations.set(productKey, qty);System.out.println(Thread.currentThread().getName() + " 库存充足");} else {System.out.println(Thread.currentThread().getName() + " 库存不足");}});}//防止主线程停止,导致模拟线程不执行Thread.sleep(60 * 1000);}

出现问题,明明库存只有5个,10个用户却都提示库存充足

考虑高并发的情况下 

使用decrement来保证原子性,解决高并发问题

    @Testvoid test3() throws InterruptedException {//获取线程池ExecutorService executorService = Executors.newCachedThreadPool();//模拟开启10个线程for (int i = 1; i <= 10; i++) {//开启线程执行executorService.execute(() -> {//获取库存Long qty = valueOperations.decrement(productKey, 1);if (qty >= 0) {System.out.println(Thread.currentThread().getName() + " 库存充足");} else {System.out.println(Thread.currentThread().getName() + " 库存不足");}});}//防止主线程停止,导致模拟线程不执行Thread.sleep(60 * 1000);}

解决了并发问题但是出现了新的问题,库存量为负数不合理(意思就是某个线程在执行删减库存时,其他线程也在同一时刻去执行,没有保障redis原子性问题)

有一种解决办法就是通过加锁synchronized来解决,但是响应能力就会大幅降低,也就失去了redis高性能的意义

所以最好的解决方式就是通过EVAL命令执行Lua脚本来解决这类问题

💌模拟多线程并发获取分布式锁SetNX

模拟某一个产品的出入库单据只能一个人(线程)操作的场景(分布式锁),当有人操作时,其他人只能等待该人操作结束之后进行操作

@SpringBootTest
@Slf4j
class AppTests_SetNx {@Resource(name = "redisTemplate")private ValueOperations valueOperations;@Autowiredprivate StringRedisTemplate stringRedisTemplate;//定义一个产品private final static String productKey = "product01";//定义锁private final static String lockKey = "lock.1";//业务private void doing(){while (true){//设置锁Boolean b = valueOperations.setIfAbsent(lockKey, 1);if(b){log.info(Thread.currentThread().getName() + " 获取到分布式锁");//业务代码开始ThreadUtil.sleep(3000);//模拟执行业务操作用时//业务代码结束stringRedisTemplate.delete(lockKey);log.info(Thread.currentThread().getName() + " 释放分布式锁");break;} else {//log.info(Thread.currentThread().getName() + " 没有获取到分布式锁,开始睡眠");ThreadUtil.sleep(2000);}}}@Testvoid test1() {//设置初始化库存5个valueOperations.set(productKey, 0);}@Testvoid test2() throws InterruptedException {//获取线程池ExecutorService executorService = Executors.newCachedThreadPool();//模拟开启10个线程for (int i = 1; i <= 10; i++) {//开启线程执行executorService.execute(this::doing);}//防止主线程停止,导致模拟线程不执行Thread.sleep(60 * 1000);}}

使用分布式锁模拟用户下单操作,防止恶意并发(一般来说用户下单产品操作时间最少也需要几秒的时间,这里预设5秒),也就是说5秒内只允许下单成功一次,也就解决了恶意并发的问题

    @Testvoid test3() throws InterruptedException {String user = "用户1";//获取线程池ExecutorService executorService = Executors.newCachedThreadPool();//模拟开启10个线程for (int i = 1; i <= 10; i++) {//开启线程执行executorService.execute(() -> {Boolean b = valueOperations.setIfAbsent("lock." + user, 1, 5, TimeUnit.SECONDS);if (b) {//下单log.info("{} 下单成功", Thread.currentThread().getName());} else {log.info("{} 稍后再试", Thread.currentThread().getName());}});}//防止主线程停止,导致模拟线程不执行Thread.sleep(60 * 1000);}

💌模拟多线程并发访问

在网关设置防止恶意并发访问,将用户ip作为key,只允许用户ip在短时间内访问2次

package com.moming;import cn.hutool.core.date.DateUtil;
import cn.hutool.core.thread.ThreadUtil;
import lombok.extern.slf4j.Slf4j;
import org.junit.jupiter.api.Test;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.test.context.SpringBootTest;
import org.springframework.data.redis.core.StringRedisTemplate;
import org.springframework.data.redis.core.ValueOperations;import javax.annotation.Resource;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;@SpringBootTest
@Slf4j
class AppTests_Increment {@Resource(name = "redisTemplate")private ValueOperations valueOperations;@Autowiredprivate StringRedisTemplate stringRedisTemplate;@Testvoid test4() throws InterruptedException {//模拟用户ipString userIp = "127.0.0.1";//获取线程池ExecutorService executorService = Executors.newCachedThreadPool();//模拟开启10个线程for (int i = 1; i <= 10; i++) {//开启线程执行executorService.execute(() -> {String seconds = DateUtil.format(DateUtil.date(), "yyyyMMddHHmmss");String key = seconds + userIp;Long count = valueOperations.increment(key, 1);if(count>2){//下单log.info("{} 超过请求次数",Thread.currentThread().getName());} else  {log.info("{} 请求正常",Thread.currentThread().getName());}});}//防止主线程停止,导致模拟线程不执行Thread.sleep(60 * 1000);}}

💌BitMap模拟在线统计 

模拟某线上活动要举办3天,参与活动的人员可以线上签到,后台通过3天人员的签到情况,统计信息

1. 3天满勤人数: 三天都签到才算一位
2. 3天活跃人数: 任何一天签到都算一位

@SpringBootTest
class AppTests_BitMap {@Autowiredprivate StringRedisTemplate stringRedisTemplate;@Testpublic void test2() {//2022111 20221112 20221113 三天活动//张三 1, 李四 2, 王五:3//满勤人数,1人//活跃人数,3人stringRedisTemplate.opsForValue().setBit("20221111",1,true);stringRedisTemplate.opsForValue().setBit("20221112",2,true);stringRedisTemplate.opsForValue().setBit("20221112",2,true);stringRedisTemplate.opsForValue().setBit("20221111",3,true);stringRedisTemplate.opsForValue().setBit("20221112",3,true);stringRedisTemplate.opsForValue().setBit("20221113",3,true);//RedisStringCommands.BitOperation.AND 满勤人数//RedisStringCommands.BitOperation.OR 活跃人数RedisCallback callback1 = connection -> {connection.bitOp(RedisStringCommands.BitOperation.OR,"人数统计".getBytes(StandardCharsets.UTF_8),"20221111".getBytes(StandardCharsets.UTF_8),"20221112".getBytes(StandardCharsets.UTF_8),"20221113".getBytes(StandardCharsets.UTF_8));Long count = connection.bitCount("人数统计".getBytes(StandardCharsets.UTF_8));return count;};Long count1 = stringRedisTemplate.execute(callback1);System.out.println(count1);}
}

RedisStringCommands.BitOperation.AND 满勤人数

RedisStringCommands.BitOperation.OR 活跃人数