Home > Article > Backend Development > Introduction to fastcgi_finish_request in php and its non-blocking code
This article brings you an introduction to fastcgi_finish_request in PHP and its non-blocking code. It has certain reference value. Friends in need can refer to it. I hope it will be helpful to you.
In actual projects, there is often such a demand. Requests sent from the front end need to be processed for a long time in the back end, but in order to provide users with better Experience, in order to prevent PHP from blocking when processing long-term tasks on the backend and quickly respond to page requests, the application of fastcgi_finish_request is summarized here. Of course, there are many ways to implement non-blocking in PHP, such as asynchronous scripts and swoole, but personally I think fastcgi_finish_request is the simplest and most convenient.
(PHP 5 >= 5.3.3, PHP 7)
fastcgi_finish_request — flush all responses Data to client
boolean fastcgi_finish_request (void)This function flushes all response data to the client and ends the request. This allows tasks that take a long time to run to continue running after the client ends the connection.
Returns TRUE on success, or returns FALSE on failure
PHP and Web server use If PHP-FPM (FastCGI Process Manager) is installed, the session can be ended immediately through the fastcgi_finish_request() function, and the PHP thread can continue to run in the background. That is to say, this function can only be used for the process management method of php-fpm
As long as the code runs to this location, the request has been disconnected and the parameters have been returned to the client. The following code has nothing to do with the client. That is to say, the content output on the page must be placed before the fastcgi_finish_request function
After fastcgi_finish_request() ends the client connection, the running time will still be affected by the max_execution_time timeout, that is to say If it is expected that the code will take a long time to execute on the backend, you still need to set set_time_limit(0)
. If the execution time is too long under high concurrency, the fastcgi process will not be used enough and cannot be released in time. , a 502 error will occur.
echo "program start..."; file_put_contents('/tmp/garylog.log','start-time:'.date('Y-m-d H:i:s')."\n", FILE_APPEND); fastcgi_finish_request();sleep(1); // set_time_limit(0); // sleep(150); $num = 25; $num += 1; sleep(5); echo 'debug...'; file_put_contents('/tmp/garylog.log', 'start-proceed:'.$num.',时间'.date('Y-m-d H:i:s')."\n", FILE_APPEND); sleep(10); file_put_contents('/tmp/garylog.log', 'end-time:'.date('Y-m-d H:i:s')."\n", FILE_APPEND);
Run test
Compatible with non-php-fpm
In terms of code portability, you can attach the following code to the code:
if (!function_exists("fastcgi_finish_request")) { function fastcgi_finish_request() { } }
It will not cause code deployment to cause problems in non-fpm environments.
Regarding the problem mentioned above: Excessive execution time under high concurrency will also cause fastcgi The process is not enough and cannot be released in time. At the same time, our requirement is only for triggering and does not need to be run every time. Then we can consider using the following method to avoid duplicating the process.
$processId = realpath(__FILE__) . '-' . get_class($this); $filename = md5($processId); $file = '/tmp/'.$filename; if(!file_exists($filename)){ file_put_contents($file, getmypid()); }else{ return true; } ## do somthing 需要长时间处理的代码 //处理完成后删除进程id记录文件 unlink($file);
Related recommendations:
The magical fastcgi_finish_request in php
##PHP uses the fastcgi_finish_request() function to implement asynchronous operations and improve responding speed
The above is the detailed content of Introduction to fastcgi_finish_request in php and its non-blocking code. For more information, please follow other related articles on the PHP Chinese website!