Apache mod_status, Nginx stub_status & PHP-FPM Status

Apache, Nginx and PHP-FPM each include a lightweight status page that shows, in real time, how many connections and workers are busy. When a site slows down under load, these pages answer the key question in seconds: is the web server out of workers, is PHP out of children, or is the bottleneck elsewhere?

Before you start

Root access to a VPS running Apache or Nginx with PHP-FPM.
The status pages reveal internal details, so they must be restricted to localhost or your IP.

Nginx stub_status

Check the module is built in: nginx -V 2>&1 | grep -o with-http_stub_status_module.

Add a local-only server or location:

server {
    listen 127.0.0.1:8081;
    location = /nginx_status { stub_status; }
}

Reload and read it: curl -s http://127.0.0.1:8081/nginx_status

Active connections: 291
server accepts handled requests
 16630948 16630948 31070465
Reading: 6 Writing: 179 Waiting: 106

Active connections: all open client connections.
accepts / handled: should be equal. If handled is lower, Nginx hit worker_connections or a file limit.
requests: total requests; divide the change by time for requests per second.
Reading / Writing / Waiting: reading request headers, sending responses (or waiting on PHP), idle keep-alive. High Writing with low traffic often means a slow backend.

Apache mod_status

Enable the module: sudo a2enmod status (on AlmaLinux it is loaded by default).

Restrict it, e.g. in /etc/apache2/mods-enabled/status.conf:

<Location "/server-status">
    SetHandler server-status
    Require local
</Location>
ExtendedStatus On

Restart, then read it: curl -s http://127.0.0.1/server-status?auto

Key fields: BusyWorkers and IdleWorkers; when BusyWorkers reaches MaxRequestWorkers, new requests queue. The Scoreboard line shows one character per slot: _ waiting, R reading, W sending reply, K keep-alive, . open slot. A scoreboard full of W means workers are stuck waiting on slow PHP or a slow client.

If a front controller rewrites every URL to index.php, exclude /server-status from the rewrite.

PHP-FPM status page

In the pool file (e.g. /etc/php/8.3/fpm/pool.d/www.conf) set:
```
pm.status_path = /fpm-status
ping.path = /fpm-ping
```

Expose it locally through Nginx:

location ~ ^/(fpm-status|fpm-ping)$ {
    allow 127.0.0.1; deny all;
    include fastcgi_params;
    fastcgi_param SCRIPT_NAME $uri;
    fastcgi_param SCRIPT_FILENAME $uri;
    fastcgi_pass unix:/run/php/php8.3-fpm.sock;
}

Restart PHP-FPM, reload Nginx, then: curl -s 'http://127.0.0.1/fpm-status?full'

active processes / max active processes: peak concurrency since start.
max children reached: how many times the pool hit pm.max_children. Any value above 0 means requests had to wait.
listen queue: requests waiting for a free child right now. Should be 0.
slow requests: requests that exceeded request_slowlog_timeout.

Reading the three together

Symptom	Likely cause	Action
FPM listen queue > 0, max children reached rising	PHP pool too small or code too slow	Check slow log; raise `pm.max_children` if RAM allows
Apache BusyWorkers = MaxRequestWorkers	Web server worker limit	Switch to event MPM, raise limit, cache more
Nginx Writing high, FPM idle	Slow clients or large downloads	Normal; check bandwidth
All idle but site slow	Database, external API or disk	Check MySQL and I/O

Graph the metrics

Monitoring tools such as Netdata, Prometheus exporters and Zabbix read these pages automatically and keep history. See install Netdata.

FAQ: web server status pages

Is it safe to leave status pages enabled?

Yes, if they are restricted to localhost or trusted IPs.

Does LiteSpeed have a status page?

Yes, real-time stats in the WebAdmin console: LiteSpeed logs and real-time stats.

Do status pages affect performance?

No measurable impact.

Next steps

Was this guide helpful?

Still stuck?

Ask MdmHost AI for a quick answer, or open a ticket and our team will take a look. We are here 24/7.

Ask MdmHost AI Open a ticket