Websites hosted on Plesk server are unavailable with 502 Bad Gateway: upstream sent too big header while reading response header

Applicable to:

Websites are unavailable:

PLESK_INFO: 502 Bad Gateway
In the log files /var/log/nginx/error.log or /var/www/vhosts/example.com/logs/proxy_error_log, the following error message can be found:

CONFIG_TEXT: [error] 14790#0: *188 upstream sent too big header while reading response header from upstream, client: 203.0.113.2, server: subdomain.example.com, request: "POST /admin_catalog.php HTTP/1.1", upstream: "http://10.0.0.1:7080/admin_catalog.php", host: "subdomain.example.com", referrer: "http://subdomain.example.com/admin_catalog.php"
Also, the following error message can be found in the log file /var/www/vhosts/example.com/logs/proxy_error_log:

CONFIG_TEXT: [error] 707#0: *2016153 FastCGI sent in stderr: "PHP message: Error checking in buffer: The buffer you checked in was not checked out" while reading upstream, client: 203.0.113.2, server: example.com, request: "POST /wp-admin/admin-ajax.php HTTP/1.1", upstream: "fastcgi://unix:///var/www/vhosts/system/example.com/php-fpm.sock:", host:"example.com", referrer: "https://example.com/wp-admin/post-new.php?wp-post-new-reload=true"

A request cannot be processed by the proxy server because of insufficient buffer size.

For one or few domains

Log into Plesk.
Go to Domains > example.com > PHP Settings and check the selected PHP handler (the field run PHP as).
Go to Domains > example.com > Apache & nginx Settings, scroll down to the field Additional nginx directives and set the following parameters in accordance with the used PHP handler:
- For PHP handler FPM application served by Nginx:
  
  CONFIG_TEXT: fastcgi_buffers 8 16k;
  fastcgi_buffer_size 32k;
- For other PHP handlers:
  
  CONFIG_TEXT: proxy_buffers 8 16k;
  proxy_buffer_size 32k;

Increase values of these parameters if it is needed.

For all domains

Increase values of these parameters if it is needed.