Skip to content
Huginn ProxyHuginn Proxy

Observability

A pre-built Prometheus + Grafana stack ships with the repo. Grafana is pre-provisioned with a data source and a dashboard. No manual import or configuration needed.

Dashboard JSON: examples/grafana/dashboards/huginn-proxy.json.

Huginn Proxy Grafana dashboard

Prerequisites

Section titled “Prerequisites”

The main proxy stack must already be running and exposing ports on the host:

  • 9090: proxy metrics (/metrics, /health, /ready, /live)
  • 9091: eBPF agent metrics (only when using the TCP SYN stack)

Prometheus reaches both via host.docker.internal from inside the observability containers.

Not running the main stack yet? Start here: Containers.

Start the observability stack

Section titled “Start the observability stack”

Run from the repo root (not from examples/):

Terminal window
docker compose -f examples/docker-compose.observability.yml up -d

This starts Prometheus and Grafana as a separate Compose project. They scrape the proxy and the eBPF agent automatically.

Open Grafana

Section titled “Open Grafana”
  1. Open http://localhost:3000 in your browser.
  2. Log in with admin / huginn.
  3. The Huginn Proxy dashboard loads as the default home dashboard.

What the dashboard covers

Section titled “What the dashboard covers”
Panel What you see
Overview Request rate, active connections, error rate (client view), P95 latency, bandwidth
TLS Handshake rate, TLS version and cipher distribution, error rate, P95 handshake duration
Fingerprinting JA4 and HTTP/2 (Akamai) extraction rate, failure rate, P95 extraction duration
Rate limiting Evaluation rate, rejection rate, breakdowns by strategy and route
Backends Request rate, error rate, P95 latency, throughput per backend, health probe results
Config hot reload Reload success/error rate, time since last successful reload, active config hash
eBPF agent Agent up, TCP SYN captured / insert failures / malformed (when eBPF stack is running)

Full PromQL for each panel: TELEMETRY.md: Grafana Dashboard Suggestions.

Prometheus scrape config (reference)

Section titled “Prometheus scrape config (reference)”

The observability Compose file already wires this up. For custom deployments:

scrape_configs:
 - job_name: huginn-proxy
 static_configs:
 - targets: ["localhost:9090"]
 scrape_interval: 15s


 - job_name: huginn-ebpf-agent
 static_configs:
 - targets: ["localhost:9091"]
 scrape_interval: 15s

Adjust hosts to your Docker/Kubernetes networking. See Telemetry for the full metric reference.

Section titled “Related”