Observability with Prometheus and Grafana

Node Exporter: Metrics and Collectors

To see the list of metrics that the Node Exporter exposes, you can execute the following command from the Prometheus server (monitoring server):

curl -s http://$server1:9100/metrics | \
grep "^# HELP"

The Node Exporter exposes a large number of metrics (285). You can count them by executing the following command:

curl -s http://$server1:9100/metrics | \
grep "^# HELP" | \
wc -l

The Node Exporter has a system called "collectors" that collect metrics from different sources. You can enable or disable these collectors using flags. To see a list of available flags and options, you can run the following command on either server1 or server2:

node_exporter --help

ℹ️ Collectors are not an exclusive feature of the Node Exporter. They are a common feature in other exporters as well.

Some collectors are disabled by default. As an example:

• filesystem and diskstats collectors are enabled by default.
• wifi collector is disabled by default.

To disable a collector, you can use the --no-collector. flag. For example, to disable the diskstats collector, you can use the following command in the Systemd entry instead of the default one (ExecStart=/usr/local/bin/node_exporter):

/usr/local/bin/node_exporter \
--no-collector.diskstats

To enable a collector, you can use the --collector. flag. For example, to enable the wifi collector, add the appropriate flag to the Systemd entry: