Node Exporter Guide: Host Metrics for Prometheus¶

Node Exporter exposes Linux host metrics in Prometheus format. It answers questions about CPU, memory, disks, filesystems, network interfaces, and the operating system; it does not replace application instrumentation or container-level metrics from cAdvisor.

Start Node Exporter in Docker¶

node-exporter:
  image: prom/node-exporter:latest
  command:
    - --path.rootfs=/host
  pid: host
  volumes:
    - /:/host:ro,rslave

For a direct Linux installation, run the released binary as a dedicated non-login user and expose port 9100 only to Prometheus or a protected monitoring network.

Add the Prometheus Job¶

scrape_configs:
  - job_name: node
    static_configs:
      - targets: ["node-exporter:9100"]

After reloading Prometheus, open Status → Targets and confirm the target is UP. A target being reachable does not mean the host is healthy; it only proves Prometheus can scrape it.

Core Queries¶

# CPU busy percentage, averaged across CPU cores
100 - (avg by (instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

# Memory available percentage
100 * node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes

# Filesystem free percentage, excluding virtual filesystems
100 * node_filesystem_avail_bytes{fstype!~"tmpfs|overlay"}
  / node_filesystem_size_bytes{fstype!~"tmpfs|overlay"}

# Network receive rate in bytes per second
sum by (instance) (rate(node_network_receive_bytes_total{device!="lo"}[5m]))

Practical Alerts¶

Start with these conditions, tune them to the workload, and include instance, mountpoint, and an owner in the alert labels or annotations.

Host down: up{job="node"} == 0 for 5 minutes.
Disk nearly full: free space below 10% and predicted to fill soon.
High CPU: sustained busy CPU, not a momentary burst.
Memory pressure: low available memory combined with swapping or OOM events.

Common Pitfalls¶

Do not alert on every mount; exclude temporary and container filesystems.
Account for ephemeral nodes and autoscaling when defining instance-based alerts.
Do not expose the exporter openly on the internet.
Use cAdvisor or Kubernetes metrics for container and pod detail; Node Exporter only sees the host view.

Quick Check¶

curl http://localhost:9100/metrics | head
curl 'http://localhost:9090/api/v1/query?query=up%7Bjob%3D%22node%22%7D'