This prometheus exporter is used to serve metrics for ha clusters and their components, as for single components. It should run inside a node of the cluster or both.
You can find the RPM pkgs for the exporter here: https://build.opensuse.org/package/show/server:monitoring/prometheus-ha_cluster_exporter.
Once installed run the exporter inside a cluster node with:
systemctl start prometheus-ha_cluster_exporter. By default it will show on http://YOUR_HOST_IP:9002/metrics.
If you open a web-browser it will serve the metrics. The exporter can't work outside a HA cluster node.
Hint: For a terraform deployment you can also read: https://github.com/SUSE/ha-sap-terraform-deployments
-
show cluster node and resource metrics via
crm_mon(pacemaker data xml) -
show corosync metrics (ring errors, quorum metrics)
-
show SBD disk health metrics
-
show DRBD metrics (local and remote disks resource metrics)
We mantain a complete list of the metric specification, usage and possible values.
Build the binary with make and run it inside a node of the ha cluster, it will show the metrics on port 9002 by default.
Use ha_cluster_exporter -h for options.
For the technical design of the exporter have look at design (this is focused on cluster_metrics).