Swoole uses task function in PHP-fpm/apache

● Create a new RedisServer.php

● The code is as follows

set(array( 'task_worker_num' => 32, 'worker_num' => 1, 'task_enable_coroutine' => true, 'heartbeat_check_interval' => 5, 'heartbeat_idle_time' => 10, )); $server->setHandler('LPUSH', function ($fd, $data) use ($server) { $taskId = $server->task($data); if ($taskId === false) { $server->send($fd, Server::format(Server::ERROR)); } else { $server->send($fd, Server::format(Server::INT, $taskId)); } }); $server->on('Finish', function($serv, $taskID, $data) { $stats = $serv->stats(); if ($stats['tasking_num'] > 10) { //tasking_num 当前正在排队的任务数 echo "剩余任务信息：" . json_encode($serv->stats()) . "\n"; } }); $server->on('Task', function ($serv, $data) { go(function () { usleep(50000); }); var_dump($data); }); $server->start();

task inside usleep (50000); Simulate task execution time

● Create a new Queue.php

● The code is as follows

connect('127.0.0.1', 9501); $x=1; while($x <= 1000) { $redis->lpush("myqueue", json_encode(array("hello".$x, "swoole"))); $x++; }

Simulate 1000 task delivery

After testing, it will be processed in 1 second

You can follow Task The speed of task execution is adjusted. task_worker_num controls the number of started processes.

● These processes are managed by the swoole bottom layer. The bottom layer will re-create new tasks after a fatal error occurs or the process exits. Process

task_worker_num

● The maximum value shall not exceed SWOOLE_CPU_NUM * 1000

● The processing time of a single task, such as 100ms, which one The process can process 1/0.1=10 tasks in 1 second

● Task delivery speed, for example, 2000 tasks

● 2000/10=200 are generated per second, you need to set task_worker_num => ; 200, enable 200 task processes

Related recommendations: [PHP Tutorial]

The above is the detailed content of Swoole uses task function in PHP-fpm/apache. For more information, please follow other related articles on the PHP Chinese website!