基于Redis有序聚集实现滑动窗口限流 - ToB企服应用市场:ToB评测及商务社交产业平台

ZADD 资源标识时间戳请求标识

复制代码

public boolean isAllow(String key) {
ZSetOperations<String, String> zSetOperations = stringRedisTemplate.opsForZSet();
// 获取当前时间戳
long currentTime = System.currentTimeMillis();
// 当前时间 - 窗口大小 = 窗口开始时间
long windowStart = currentTime - period;
// 删除窗口开始时间之前的所有数据
zSetOperations.removeRangeByScore(key, 0, windowStart);
// 统计窗口中请求数量
Long count = zSetOperations.zCard(key);
// 如果窗口中已经请求的数量超过阈值，则直接拒绝
if (count >= threshold) {
return false;
}
// 没有超过阈值，则加入集合
String value = "请求唯一标识（比如：请求流水号、哈希值、MD5值等）";
zSetOperations.add(key, String.valueOf(currentTime), currentTime);
// 设置一个过期时间，及时清理冷数据
stringRedisTemplate.expire(key, period, TimeUnit.MILLISECONDS);
// 通过
return true;
}

复制代码

local key = KEYS[1]
local current_time = tonumber(ARGV[1])
local window_size = tonumber(ARGV[2])
local threshold = tonumber(ARGV[3])
redis.call('ZREMRANGEBYSCORE', key, 0, current_time - window_size)
local count = redis.call('ZCARD', key)
if count >= threshold then
return tostring(0)
else
redis.call('ZADD', key, tostring(current_time), current_time)
return tostring(1)
end

复制代码

package com.example.demo.controller;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.data.redis.core.StringRedisTemplate;
import org.springframework.data.redis.core.ZSetOperations;
import org.springframework.data.redis.core.script.DefaultRedisScript;
import org.springframework.stereotype.Service;
import java.util.Collections;
import java.util.concurrent.TimeUnit;
/**
* 基于Redis有序集合实现滑动窗口限流
* @Author: ChengJianSheng
* @Date: 2024/12/26
*/
@Service
public class SlidingWindowRatelimiter {
private long period = 60*1000; // 1分钟
private int threshold = 3; // 3次
@Autowired
private StringRedisTemplate stringRedisTemplate;
/**
* RedisTemplate
*/
public boolean isAllow(String key) {
ZSetOperations<String, String> zSetOperations = stringRedisTemplate.opsForZSet();
// 获取当前时间戳
long currentTime = System.currentTimeMillis();
// 当前时间 - 窗口大小 = 窗口开始时间
long windowStart = currentTime - period;
// 删除窗口开始时间之前的所有数据
zSetOperations.removeRangeByScore(key, 0, windowStart);
// 统计窗口中请求数量
Long count = zSetOperations.zCard(key);
// 如果窗口中已经请求的数量超过阈值，则直接拒绝
if (count >= threshold) {
return false;
}
// 没有超过阈值，则加入集合
String value = "请求唯一标识（比如：请求流水号、哈希值、MD5值等）";
zSetOperations.add(key, String.valueOf(currentTime), currentTime);
// 设置一个过期时间，及时清理冷数据
stringRedisTemplate.expire(key, period, TimeUnit.MILLISECONDS);
// 通过
return true;
}
/**
* Lua脚本
*/
public boolean isAllow2(String key) {
String luaScript = "local key = KEYS[1]\n" +
"local current_time = tonumber(ARGV[1])\n" +
"local window_size = tonumber(ARGV[2])\n" +
"local threshold = tonumber(ARGV[3])\n" +
"redis.call('ZREMRANGEBYSCORE', key, 0, current_time - window_size)\n" +
"local count = redis.call('ZCARD', key)\n" +
"if count >= threshold then\n" +
" return tostring(0)\n" +
"else\n" +
" redis.call('ZADD', key, tostring(current_time), current_time)\n" +
" return tostring(1)\n" +
"end";
long currentTime = System.currentTimeMillis();
DefaultRedisScript<String> redisScript = new DefaultRedisScript<>(luaScript, String.class);
String result = stringRedisTemplate.execute(redisScript, Collections.singletonList(key), String.valueOf(currentTime), String.valueOf(period), String.valueOf(threshold));
// 返回1表示通过，返回0表示拒绝
return "1".equals(result);
}
}

复制代码

String requestId = UUID.randomUUID().toString();
DefaultRedisScript<String> redisScript = new DefaultRedisScript<>(luaScript, String.class);
String result = stringRedisTemplate.execute(redisScript,
Collections.singletonList(key),
requestId,
String.valueOf(period),
String.valueOf(threshold));

复制代码

@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.METHOD)
public @interface MyAnnotation {
String value();
}
@Aspect
@Component
public class MyAspect {
@Before("@annotation(myAnnotation)")
public void beforeMethod(JoinPoint joinPoint, MyAnnotation myAnnotation) {
// 获取注解参数
String value = myAnnotation.value();
System.out.println("Annotation value: " + value);
// 其他业务逻辑...
}
}

复制代码

package com.example.demo.controller;
import java.lang.annotation.*;
/**
* 请求速率限制
* @Author: ChengJianSheng
* @Date: 2024/12/26
*/
@Documented
@Target(ElementType.METHOD)
@Retention(RetentionPolicy.RUNTIME)
public @interface RateLimit {
/**
* 窗口大小（默认：60秒）
*/
long period() default 60;
/**
* 阈值（默认：3次）
*/
long threshold() default 3;
}

复制代码

package com.example.demo.controller;
import jakarta.servlet.http.HttpServletRequest;
import lombok.extern.slf4j.Slf4j;
import org.aspectj.lang.JoinPoint;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Before;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.data.redis.core.StringRedisTemplate;
import org.springframework.data.redis.core.ZSetOperations;
import org.springframework.stereotype.Component;
import org.springframework.web.context.request.RequestContextHolder;
import org.springframework.web.context.request.ServletRequestAttributes;
import org.springframework.web.servlet.support.RequestContextUtils;
import java.util.concurrent.TimeUnit;
/**
* @Author: ChengJianSheng
* @Date: 2024/12/26
*/
@Slf4j
@Aspect
@Component
public class RateLimitAspect {
@Autowired
private StringRedisTemplate stringRedisTemplate;
// @Autowired
// private SlidingWindowRatelimiter slidingWindowRatelimiter;
@Before("@annotation(rateLimit)")
public void doBefore(JoinPoint joinPoint, RateLimit rateLimit) {
// 获取注解参数
long period = rateLimit.period();
long threshold = rateLimit.threshold();
// 获取请求信息
ServletRequestAttributes servletRequestAttributes = (ServletRequestAttributes) RequestContextHolder.getRequestAttributes();
HttpServletRequest httpServletRequest = servletRequestAttributes.getRequest();
String uri = httpServletRequest.getRequestURI();
Long userId = 123L; // 模拟获取用户ID
String key = "limit:" + userId + ":" + uri;
/*
if (!slidingWindowRatelimiter.isAllow2(key)) {
log.warn("请求超过速率限制！userId={}, uri={}", userId, uri);
throw new RuntimeException("请求过于频繁！");
}*/
ZSetOperations<String, String> zSetOperations = stringRedisTemplate.opsForZSet();
// 获取当前时间戳
long currentTime = System.currentTimeMillis();
// 当前时间 - 窗口大小 = 窗口开始时间
long windowStart = currentTime - period * 1000;
// 删除窗口开始时间之前的所有数据
zSetOperations.removeRangeByScore(key, 0, windowStart);
// 统计窗口中请求数量
Long count = zSetOperations.zCard(key);
// 如果窗口中已经请求的数量超过阈值，则直接拒绝
if (count < threshold) {
// 没有超过阈值，则加入集合
zSetOperations.add(key, String.valueOf(currentTime), currentTime);
// 设置一个过期时间，及时清理冷数据
stringRedisTemplate.expire(key, period, TimeUnit.SECONDS);
} else {
throw new RuntimeException("请求过于频繁！");
}
}
}

复制代码

@RestController
@RequestMapping("/hello")
public class HelloController {
@RateLimit(period = 30, threshold = 2)
@GetMapping("/sayHi")
public void sayHi() {
}
}

复制代码