Example configurations

Replacing Munin

This example shows how to use Duct to replace a Munin polling setup by collecting metrics from munin-node and forwarding them to a Riemann → InfluxDB → Grafana stack (DRIG).

Step 1: Stand up the DRIG stack with Docker

The easiest way to run Riemann, InfluxDB and Grafana together is with Docker Compose. Create a docker-compose.yml file:

services:
  riemann:
    image: riemannio/riemann:latest
    ports:
      - "5555:5555"
      - "5555:5555/udp"
    volumes:
      - ./riemann.config:/etc/riemann/riemann.config

  influxdb:
    image: influxdb:1.8
    ports:
      - "8086:8086"
    environment:
      INFLUXDB_DB: riemann

  grafana:
    image: grafana/grafana:latest
    ports:
      - "3000:3000"
    environment:
      GF_SECURITY_ADMIN_PASSWORD: admin

Then create a minimal riemann.config that writes events to InfluxDB:

(logging/init {:file "/var/log/riemann/riemann.log"})

(let [host "0.0.0.0"]
  (tcp-server {:host host})
  (udp-server {:host host}))

(periodically-expire 60)

(def influx
  (influxdb {:host "influxdb"
             :port 8086
             :db   "riemann"
             :version :0.9}))

(streams
  (async-queue! :influx {:queue-size 1000}
    influx))

Start the stack:

docker compose up -d

Step 2: Install and configure Duct

Install Duct:

pip install ducted

Create /etc/duct/duct.yml:

ttl: 60.0
interval: 1.0

outputs:
  - output: duct.outputs.riemann.RiemannTCP
    server: localhost
    port: 5555

sources:
  - service: mymunin
    source: duct.sources.munin.MuninNode
    interval: 60.0
    ttl: 120.0
    critical:
      mymunin.system.load.load: "> 2"

This configures Duct to connect to munin-node on the local machine and collect all configured plugin values every 60 seconds.

Start Duct:

ductd -c /etc/duct/duct.yml

Step 3: Add dashboards in Grafana

Open Grafana at http://localhost:3000 (default login: admin / admin).

1. Add an InfluxDB data source pointing at http://influxdb:8086, database riemann.

2. Create a new dashboard and add panels querying the metric series written by Riemann. Series names follow the pattern <host>.<service>, for example myhost.mymunin.network.if_eth0.down.

Many Munin metrics are counter types. The duct.sources.munin.MuninNode source handles this automatically by caching the previous value and emitting rates of change.

Using Prometheus instead of Riemann

If you prefer a Prometheus pull model, Duct can expose a scrape endpoint directly. Replace the Riemann output with:

outputs:
  - output: duct.outputs.prometheus.Prometheus
    port: 9110
    prefix: duct_

Then point your Prometheus scrape config at http://<host>:9110/metrics. No separate time-series database or Riemann instance is needed.