Redis actual combat limit operation frequency

I have been obsessed with business development recently and haven’t updated my blog for a while. I plan to include some practical solutions in business scenarios or better design ideas in subsequent blog posts. Sharing is not like spending a lot of time sorting out related content around a topic before (holding back the ultimate trick). The content of the subsequent article may not be so rich, but it can be more detailed on one point as much as possible, or More in-depth analysis, through continuous sharing and self-review, accumulation of experience, while increasing the frequency of blog sharing

SCENARIO

SCENARIO 1

留言功能限制，30秒 内只能评论 10次，超出次数不让能再评论，并提示：过于频繁

SCENARIO 2

点赞功能限制，10秒 内只能点赞 10次，超出次数后不能再点赞，并禁止操作 1个小时，提示：过于频繁，被禁止操作1小时

Scenario 3

上传记录功能，限制一天只能上传 100次，超出次数不让能再上传，并提示：超出今日上线

Abstracting from the essence

In the process of business development, we are constantly involved in the design of various business scenarios, and it is often easy to encounter very similar problems Scenarios, but the business modules they currently belong to are different. In fact, the essence of these requirements is to solve the same problem. When encountering this scenario, we need to extract the essential issues of the requirements based on our own experience analysis and implement a universal Solutions, make your solutions more valuable, this may be the difference between you being an engineer with a soul or the strongest cp (copy paste) king.

Analyzing the above three business scenarios, we can find that they have similar logic and call them similar problems. Now we are going to isolate this problem, design a general solution, and outline the same logic flow chart :

By analyzing the above demand scenarios, we can extract the conditions they all need:

Restricted objects: User restricted operations (comments, likes) , record, ...) within the time range

（最小时间单位用秒：天/小时/分钟都可换算成秒，用秒可以解决更多的场景）

1,'ttl'=>过期时间/秒] ['type'=>2,'ttl'=>具体过期时间戳] 二选一
 * @return bool
 * @throws \Exception
 */public static function frequencyLimit(string $action, int $userId, int $time, int $number, $expire = []){    // todo 根据用户操作动作时间范围，进行频率的控制和失效释放}

The function needs to carry out the operations and time initiated by the user, as well as the cumulative number of times Storage, and need to clean up expiration. If we rely on mysql for storage at this time, it will be quite painful to think about it. The protagonist here: redis finally appears. Based on the characteristics of redis, the atomic operation of incr and key support the expiration mechanism and memory storage. The efficiency advantage can accomplish the purpose relatively simply, flexibly and efficiently.

Here is a simple code to implement a common function:

1,'ttl'=>过期时间/秒] ['type'=>2,'ttl'=>具体过期时间戳] 二选一
 * @return bool
 * @throws \Exception
 */public function frequencyLimit(string $action, int $userId, int $time, int $number, $expire = []){    if (empty($action) || $userId <= 0 || $time <= 0 || $number <= 0) {        throw new \Exception('非法参数');
    }
    $key = 'act:limit:' . $action . ':' . $userId;
    $r = RedisClient::connect();    //获取当前累计次数
    $current = intval($r->get($key));    if ($current >= $number) return false;    //累计并返回最新值
    $current = $r->incr($key);    //第一次累加，设置控制操作频率的有效时间
    if ($current === 1) $r->expire($key, $time);    //未超出限制次数先放过
    if ($current < $number) return true;    //超出后根据需要重新设置过期失效时间 $current === $number 判断保证只重新设置一次
    $type = empty($expire['type']) ? 0 : intval($expire['type']);
    $ttl = empty($expire['ttl']) ? 0 : intval($expire['ttl']);    if ($current === $number && $ttl > 0 && in_array($type, [1, 2])) {        if ($type === 1) $r->expire($key, $ttl);        if ($type === 2) $r->expireAt($key, $ttl);
    }    return false;
}//场景1/**
 * 评论限制
 * @param int $userId
 * @return bool|string
 */public function doComment(int $userId){    try {
        $pass = FrequencyLimit::doHandle('comment', $userId, 30, 10);        if (!$pass) return '过于频繁';        // todo 评论逻辑
        return true;
    } catch (\Exception $e) {        return $e->getMessage();
    }
}//场景2/**
 * 点赞限制
 * @param int $userId
 * @return bool|string
 */public function doLike(int $userId){    try {
        $pass = FrequencyLimit::doHandle('like', $userId, 10, 10, ['type' => 1, 'ttl' => 1 * 60 * 60]);        if (!$pass) return '过于频繁，被禁止操作1小时';        // todo 点赞逻辑
        return true;
    } catch (\Exception $e) {        return $e->getMessage();
    }
}//场景3/**
 * 上传限制
 * @param int $userId
 * @return bool|string
 */public function doUpload(int $userId){    try {
        $expire = strtotime(date('Y-m-d', strtotime(+1 . 'days')));
        $pass = FrequencyLimit::doHandle('upload', $userId, 1 * 24 * 60 * 60, 100, ['type' => 2, 'ttl' => $expire]);        if (!$pass) return '超出今日上线';        // todo 上传逻辑
        return true;
    } catch (\Exception $e) {        return $e->getMessage();
    }
}//场景N

编码上可以根据你设计这个通用方案的复杂度进行进一步抽象，如抽象成频率限制的功能类等

Summary

Analyze similar business scenarios, discover essential problems and design general solutions

Make solutions more valuable and be a developer with a soul

Master redis proficiently and make full use of its features and advantages

For more Redis-related technical articles, please visit

column to learn!

The above is the detailed content of Redis actual combat limit operation frequency. For more information, please follow other related articles on the PHP Chinese website!