• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Issue Custom PHP FPM configuration. Configuration file disappears when registering with php_handler

RDrake

New Pleskian
I installed a custom PHP version in order to be able to use PHP parallel extension.
After the installation, the service was missing, so I set it up by copying
plesk-php82-fpm.service and the directory plesk-php82-fpm.service.d/

I then registered the new PHP version according to this article

At first it seemed to work like a charm but then I got 503 errors when visiting the website.
I started to observe the /php-fpm.d/ directory of the new php version. After assigning a domain to the new php version in plesk backend, two config files appear:
Code:
mydomain.de.conf
plesk-service.localdomain.conf

The service status is:
Code:
● php82-fpm.service - The PHP 8.2 FastCGI Process Manager
   Loaded: loaded (/lib/systemd/system/php82-fpm.service; enabled; vendor preset: enabled)
  Drop-In: /lib/systemd/system/php82-fpm.service.d
           └─limit_nofile.conf, respawn.conf
   Active: activating (start) since Sun 2023-04-09 15:36:02 CEST; 1min 17s ago
 Main PID: 21759 (php-fpm)
    Tasks: 1 (limit: 105)
   CGroup: /system.slice/php82-fpm.service
           └─21759 php-fpm: master process (/etc/php8z/etc/php-fpm.conf)

Apr 09 15:36:02 h2891136.stratoserver.net systemd[1]: Starting The PHP 8.2 FastCGI Process Manager...

But after about one minute the config files disappear and the status of the service goes to the failed state :

Code:
● php82-fpm.service - The PHP 8.2 FastCGI Process Manager
   Loaded: loaded (/lib/systemd/system/php82-fpm.service; enabled; vendor preset: enabled)
  Drop-In: /lib/systemd/system/php82-fpm.service.d
           └─limit_nofile.conf, respawn.conf
   Active: failed (Result: exit-code) since Sun 2023-04-09 15:37:34 CEST; 3min 23s ago
 Main PID: 17614 (code=exited, status=78)
[09-Apr-2023 15:37:33] ERROR: No pool defined. at least one pool section must be specified in config file


I have no idea on how I analyse or fix this.
Neither
Code:
/var/log/sw-cp-server/sw-engine.log
, nor the php82-fpm.log errorlog show a related error (except for
Code:
ERROR: No pool defined. at least one pool section must be specified in config file
)

Any hints on how to investigate this are greatly welcome!
 
I just got aware the there is a timeout after assigning the new PHP version in the backend:

Code:
● php82-fpm.service - The PHP 8.2 FastCGI Process Manager
   Loaded: loaded (/lib/systemd/system/php82-fpm.service; enabled; vendor preset: enabled)
  Drop-In: /lib/systemd/system/php82-fpm.service.d
           └─limit_nofile.conf, respawn.conf
   Active: activating (auto-restart) (Result: timeout) since Sun 2023-04-09 15:55:45 CEST; 99ms ago
  Process: 26718 ExecStart=/etc/php8z/sbin/php-fpm --nodaemonize --fpm-config /etc/php8z/etc/php-fpm.conf (code=exited, status=0/SUCCESS)
 Main PID: 26718 (code=exited, status=0/SUCCESS)

--> would be great if there was an error log which could provide more insight on this ... ??
 
I just got aware the there is a timeout after assigning the new PHP version in the backend:

Code:
● php82-fpm.service - The PHP 8.2 FastCGI Process Manager
   Loaded: loaded (/lib/systemd/system/php82-fpm.service; enabled; vendor preset: enabled)
  Drop-In: /lib/systemd/system/php82-fpm.service.d
           └─limit_nofile.conf, respawn.conf
   Active: activating (auto-restart) (Result: timeout) since Sun 2023-04-09 15:55:45 CEST; 99ms ago
  Process: 26718 ExecStart=/etc/php8z/sbin/php-fpm --nodaemonize --fpm-config /etc/php8z/etc/php-fpm.conf (code=exited, status=0/SUCCESS)
 Main PID: 26718 (code=exited, status=0/SUCCESS)

--> would be great if there was an error log which could provide more insight on this ... ??
Further info: this is not fixed by executing
plesk repair web mydomain.de
 
Also not fixed by
Code:
# plesk repair web -php-handlers -php-fpm-configuration

Another info: the webpage works like a charm with the new PHP version - exactly until the timeout happens %)
 
Whadda ya know! I stumbled across a similar error like mine:
php7.0-fpm.service: Start operation timed out. Terminating. - Raspberry Pi Forums
And so I simply installed the Debian -- Informationen über Paket rng-tools in stretch in my system and bam! it just works now.

If anyone has an explanation why all plesk-installed PHP versions just work, but the custom configured PHP version needs an extra set of tools - that doesn't get installed automatically - I'd be thankful for a small enlightenment.
 
Back
Top