How to configure Prometheus to scrape node exporter

Prometheus needs a scrape job for each node exporter endpoint before Linux host metrics appear in queries, alerts, or dashboards. Adding a node exporter target to /etc/prometheus/prometheus.yml tells Prometheus to poll the host's /metrics endpoint, store the node_* time series, and mark the target up or down from the scrape result.

Node exporter listens on HTTP port 9100 by default. A static scrape target normally needs only the hostname or IP address plus that port because Prometheus already uses /metrics as the default scrape path.

Save the new job in the active Prometheus configuration file, validate it with promtool, reload Prometheus, and query both up{job=“node”} and a node_ metric. A target that answers from a laptop can still stay down from Prometheus, so run the reachability check from the Prometheus host or from the same network path.

1. Confirm node exporter answers from the Prometheus host.

   $ curl http://node01.example.net:9100/metrics
   # HELP go_gc_duration_seconds A summary of the pause duration of garbage collection cycles.
   # TYPE go_gc_duration_seconds summary
   ##### snipped #####
   # HELP node_load1 1m load average.
   # TYPE node_load1 gauge
   node_load1 0.42

   Use the hostname, IP address, and port that Prometheus will scrape. Node exporter listens on port 9100 unless it was started with a custom web listen address.

2. Open the active Prometheus configuration file.

   $ sudo vi /etc/prometheus/prometheus.yml

   Use the path from Prometheus' --config.file flag when the service was installed outside the distro package layout.

3. Add a node exporter scrape job under scrape_configs.

   scrape_configs:
     - job_name: node
       static_configs:
         - targets:
             - node01.example.net:9100

   Keep any existing scrape_configs entries and add only the new job_name: node block when the file already contains other scrape jobs.

4. Check the Prometheus configuration syntax.

   $ promtool check config /etc/prometheus/prometheus.yml
   Checking /etc/prometheus/prometheus.yml
    SUCCESS: /etc/prometheus/prometheus.yml is valid prometheus config file syntax

5. Reload Prometheus.

   $ curl -X POST http://localhost:9090/-/reload

   A successful lifecycle reload returns an empty response body. The endpoint requires Prometheus to start with --web.enable-lifecycle; use systemctl reload prometheus or the local service manager when the endpoint is disabled.
   Related: How to reload Prometheus configuration
   Related: How to manage the Prometheus service with systemctl

6. Check the node exporter target state.

   $ promtool query instant http://localhost:9090 'up{job="node"}'
   up{instance="node01.example.net:9100", job="node"} => 1

   Value 1 means Prometheus scraped the target successfully. Value 0 means Prometheus knows the target but the latest scrape failed.
   Related: How to check Prometheus targets

7. Query a node exporter metric from Prometheus.

   $ promtool query instant http://localhost:9090 'node_load1{job="node"}'
   node_load1{instance="node01.example.net:9100", job="node"} => 0.42

   A returned node_ metric proves the target is not only listed, but also sending host metrics into Prometheus.
   Related: How to run a PromQL query in Prometheus