yggprom

Prometheus exporter for Yggdrasil peer status, traffic totals and rates, latency, cost, and uptime.

Usage

Requires Go 1.24+ and access to a running Yggdrasil node's UNIX admin socket.

go build -o yggprom .
./yggprom -name my-node -listen 127.0.0.1:9000

Serves metrics over HTTP at /.

Flag	Description	Default	Required
-name	Node label value	default	No
-listen	TCP listen address	:0 (any available port)	No
-admin	UNIX admin socket path	/var/run/yggdrasil/yggdrasil.sock	No

Metrics

Metrics are collected from Yggdrasil's getPeers admin API on each scrape. TX and RX are relative to the local node.

Metric	Meaning
yggdrasil_peer_up	Connection status: 1 connected, 0 disconnected.
yggdrasil_peer_tx_total_bytes	Total bytes sent over the current connection.
yggdrasil_peer_rx_total_bytes	Total bytes received over the current connection.
yggdrasil_peer_tx_rate_bytes	Current send rate in bytes per second, reported by Yggdrasil.
yggdrasil_peer_rx_rate_bytes	Current receive rate in bytes per second, reported by Yggdrasil.
yggdrasil_peer_latency_ms	Peer round-trip latency in milliseconds.
yggdrasil_peer_cost	Yggdrasil routing cost for the peer; lower values are preferred.
yggdrasil_peer_uptime_secs	Current connection uptime in seconds.

Connected peers have labels node (from -name), peer (URI), dir (inbound or outbound), and key (peer public key). Disconnected peers emit only yggdrasil_peer_up{peer="…"} 0. Traffic totals and uptime reset on reconnection.

Prometheus

scrape_configs:
  - job_name: yggdrasil
    metrics_path: /
    static_configs:
      - targets: ['127.0.0.1:9000']