How does Amplify measure NGINX requests per second?

The agent parses nginx_status (stub_status module) once a minute and derives the per-second rate from the accepts and requests counters. No log parsing required.

Can I alert on sudden RPS drops?

Yes. Configure alerts on nginx.http.request.count with threshold comparisons (absolute or percentage-of-baseline) and route to Slack, Telegram, email or webhooks.

Does it track per-upstream RPS?

Yes — with NGINX Plus the agent reports per-upstream request rate. On open-source NGINX, RPS is reported per server block.

NGINX Requests Per Second Monitoring
Know when traffic moves.

Real-time per-server RPS, status-code breakdown, and upstream request rate. Alerts fire on both spikes and unexpected drops.

Start free See demo

What Amplify reports for RPS

nginx.http.request.count

Total requests handled per second, derived from stub_status.

nginx.http.status.1xx / 2xx / 3xx / 4xx / 5xx

Per-status-class request rate, computed from access-log parsing.

nginx.upstream.request.count

Per-upstream request rate (NGINX Plus only).

nginx.http.method.get / post / put

Per-method request rate, computed from the access log.

Setup in 5 minutes

# Install the agent
curl -sS https://amplify.getpagespeed.com/install.sh | sudo bash

# Expose stub_status (required for RPS)
location = /nginx_status {
    stub_status;
    allow 127.0.0.1;
    deny all;
}

After restart, RPS graphs show up immediately under Host → NGINX.

Watch NGINX traffic the moment it shifts.

Start free View pricing

NGINX Requests Per Second MonitoringKnow when traffic moves.