// the find
nginx/nginx-prometheus-exporter
NGINX Prometheus Exporter for NGINX and NGINX Plus
A Go exporter that scrapes NGINX's stub_status page (open source) or the API (NGINX Plus) and translates the numbers into Prometheus metrics. It's for anyone running NGINX who wants it in a Prometheus/Grafana stack without writing a custom scraper.
Single static binary with no runtime deps, so it drops into a sidecar, systemd unit, or k8s pod trivially. For NGINX Plus it covers a lot of ground: server zones, upstreams, SSL handshakes, rate limiting, resolver stats, zone sync - not just the basics. Supports unix socket scraping, mTLS client certs to the upstream, and repeatable scrape URIs for multiple targets from one process. It's maintained by nginx/F5 directly with SBOM and provenance attestation on releases, so it's not a random community fork that dies in a year.
Open source NGINX only gets 7 metrics from stub_status - connections and total requests, nothing about latency, status codes, or per-vhost traffic. That richer data is Plus-only, which is a paywall a lot of readers will hit immediately. It's a straight metrics translator with a fixed naming scheme; there's no way to relabel or drop metrics you don't want without a Prometheus-side relabel config. It's explicitly not a daemon, so you own the restart/supervision logic yourself. And every scrape is synchronous against NGINX directly - a slow or hung stub_status/api response stalls that scrape cycle with no caching layer in between.