Hello,

I have a timeout problem in a server that is hosting several different pages (on a few different domains). For some of these pages, the execution of the PHP scripts can't be more than 45 seconds (but it is not PHP that puts this limit, for me).

To test I have made a simple script with a "sleep(46);" call. With 46 the page will be a 500 HTTP "internal server error". With 44 or less it will work, and a call to phpinfo() function in this test script shows "max_execution_time 6000", despite of the 45 second limitation.

Nginx's configuration for this site (/var/www/[site-path]/conf/last_nginx.conf) doesn't have anything to limit the execution time. The global Nginx config also doesn't have anything. I thought I could add something to Nginx config (assuming it's his fault), but on documentation pages I didn't see anything that may help. http://nginx.org/en/docs/dirindex.html] Am I right?

So I'm guessing it must be some system program or limit that imposes this limit for some sites (since not all of them have this limitation). But I need help on finding out wich programs or configuration files to check, change or create. Could you help me?

Any ideia that may be a single step to the solution is very welcome.

Maxmilhas

Is the max_execution_time in /etc/php.ini set to something sane?

You say that you are getting a 500 error, there should (hopefully) be a log somewhere that has more info. You may need to enable the "display_errors" option in php.ini.

Posting your nginx, php, and fpm configuration will help as well.

Also, I've deployed php applications on a LOT of CentOS servers and I've always used apache. I use Nginx for pretty much anything else, but Php+apache always have worked out well for me on CentOS, just thought I'd mention that if you are having issues with FPM.

Yes, it is set to 3600. So, 6000 in local php.ini (and phpinfo() output), and 3600 in global php.ini.

I'll check here (for the other parts of your answer) and reply again soon. Thanks