Simple PHP to control site concurrency_PHP tutorial

Generally speaking, there will always be some difficult business to handle in a project, such as search with complex business and various styles. If you use a search system such as sphinx, lucence, etc. to handle it, If the data source changes too quickly (such as the number of likes, dislikes, views, etc.), it will be difficult to measure the frequency of pushing the data first, and it will not be possible to search or sort accurately. Therefore, the general approach is to search through the DB, and in addition to decomposing the business to the program level as much as possible, a layer of cache will be added in front of the DB. However, there are many known disadvantages in doing so:
1. In fact, many businesses cannot be processed in ordinary PHP, such as some more complex sorting and filtering, which can be implemented through ordinary PHP The efficiency is not as good as MYSQL.
2. The cache hit rate is not guaranteed, especially for some businesses that use keywords to query, the keywords change too much.

If encountering malicious brushing, the DB server will easily hang up directly, so we can add a layer of limiter at the program level to limit the number of concurrencies. The limiter has the following characteristics:
1. High efficiency (nonsense, otherwise the limiter will be maxed out if you just brush it). Currently, memcache is used for atomic operation counting, which can be extended to use other methods.
2. Accurate to the action layer. Individual pages can be restricted individually.
3. Convenience. Different people have different opinions
4. Low cost. Including R&D costs and hardware costs.

According to the stress test, the search function was originally only able to have 100 concurrencies. After limiting it to 100, the stress test result of 800 was normal.

The code is as follows:

<?php
    /**
     * 搜索
     */
    public function search(){
        //增加计数器功能，超过次数则返回系统繁忙
        $viewlimiter = Library::load('viewlimiter');
        //访问限制器的名称，请用controller+action，尽量单个页面保持唯一
        $limitName = 'search_search';
        //最多同时访问数
        $limit = 100; 
        //若是超过数量则直接返回
        if(!$viewlimiter->check($limitName, $limit)) {
            ajaxOutput(0, $this->lang->line('multi_search_limit'));
        }


        /*----复杂变态的业务逻辑-----*/
        $xxxModel->search($params);
    }
    //end func
?>

Extended topic:
1. If the business is brushed, normal users may not be able to view the content. -- This problem can only be solved from the system level, which can be seen here: http://johnsteven.blog.51cto.com/2523007/818209
2. It is best to limit the number Combined with the cache hit rate, the limit number is automatically changed within a certain range to improve the accuracy of the limit. -- This can be studied later and combined with the cache class
3. Dynamic control based on server load. -- Involving the bottom layer and server permissions, the research cost is relatively high. If you are interested, you can research it.

Note:
This article only explains the ideas, the specific code is temporarily not available due to some reasons

http://www.bkjia.com/PHPjc/440330.htmlwww.bkjia.comtruehttp: //www.bkjia.com/PHPjc/440330.htmlTechArticleGenerally speaking, there are always some difficult businesses to handle in a project, such as complex and diverse businesses. Search, if you use a search system such as sphinx, lucence, etc. to process it, the data source...