Forum GLPI-Project

drosera

Member

Registered: 2020-06-28

Posts: 41

NGINX Gateway timeouts when deleting objects: How to troubleshoot?

Hi,

we have an issue, which has bothered us quite a while since GLPI 9.x and it hasn't gone away with the recent update to GLPI 10.0.11: Whenever we delete e.g. a PC (with all devices) from the trashbin, the webserver (NGINX) runs into a timeout ("504 Gateway Time-out"). The URL at that time is normally https://<GLPI-SERVER>/front/massiveaction.php. A reload of the page doesn't help and not even going back to the GLPI main dashboard (https://<GLPI-SERVER>/) is possible, as we still get the timeout on this page. Instead we need to delete all site cookies, reload and log in from scratch. This is of course quite cumbersome :-/. At least once we log in again, the objects have been deleted. ;-)

Our NGINX proxy configuration looks like this:

   server {
      listen       443 ssl http2;
      listen       [::]:443 ssl http2;
      server_name  <GLPI-SERVER>;
      root         /var/www/glpi;

      ssl_certificate "/etc/pki/tls/<GLPI-SERVER>.crt";
      ssl_certificate_key "/etc/pki/tls/private/<GLPI-SERVER>.key";
      ssl_session_cache shared:SSL:1m;
      ssl_session_timeout  10m;
      ssl_ciphers HIGH:!aNULL:!MD5;
      ssl_prefer_server_ciphers on;

      # Load configuration files for the default server block.
      include /etc/nginx/default.d/*.conf;

      location / {
         index index.php;
      }
   }

There is also a `php-fpm.conf`:

upstream php-fpm {
        server 127.0.0.1:9000;
        #server unix:/run/php-fpm/www.sock;
}

and `/etc/nginx/default.d/php.conf`

index index.php index.html index.htm;

location ~ \.(php|phar)(/.*)?$ {
    fastcgi_split_path_info ^(.+\.(?:php|phar))(/.*)$;

    fastcgi_intercept_errors on;
    fastcgi_index  index.php;
    include        fastcgi_params;
    fastcgi_param  SCRIPT_FILENAME  $document_root$fastcgi_script_name;
    fastcgi_param  PATH_INFO $fastcgi_path_info;
    fastcgi_pass   php-fpm;
}

As far as I remember, both are the system default and haven't been installed or ever changed by us (but I might be wrong).

The error messages we get in the NGINX error log are like (manual wrapping done by me):

2024/01/27 21:01:04 [error] 200086#200086: *62 upstream timed out (110: Connection timed out) while reading upstream,
client: 172.22.162.27, server: <GLPI-SERVER>, request: "POST /front/massiveaction.php HTTP/2.0", upstream:
"fastcgi://127.0.0.1:9000", host: "<GLPI-SERVER>", referrer: "https://<GLPI-SERVER>/front/computer.php?
is_deleted=1&as_map=0&browse=0&criteria%5B0%5D%5Blink%5D=AND&criteria%5B0%5D%5Bfield%5D=view&
criteria%5B0%5D%5Bsearchtype%5D=contains&criteria%5B0%5D%5Bvalue%5D=b260&itemtype=Computer&start=0
&_glpi_csrf_token=727af4bcf64d268f192748ca450368336da5c96a4b5be1b4d8ec2bafc6a47e8e"

The server is running in a container with 4 virtual cores and 8 GB RAM on CentOS 7.9.

I am greatefull for any hints on how to troubleshoot the problem or even the solution, if anyone has already had and solved this issue

Last edited by drosera (2024-01-28 15:25:40)

drosera

Member

Registered: 2020-06-28

Posts: 41

Re: NGINX Gateway timeouts when deleting objects: How to troubleshoot?

I have switched PHP-FPM from port to socket, but it doesn't change the situation