Configuring Prometheus to collect Kudu metrics

Learn how to configure Prometheus to collect Kudu metrics by updating the Prometheus configuration file.

Before configuring Prometheus and Grafana, ensure that you have the following components available:

A running Kudu cluster with at least one Master node and one Tablet Server node.
An installed Prometheus instance.
An installed Grafana instance.

Open the prometheus.yml configuration file in a text editor.

Add the Kudu scrape job to the scrape_configs section.

The following example shows a static target configuration:


global:
  scrape_interval: 15s
  evaluation_interval: 15s

scrape_configs:
  - job_name: "kudu"
    metrics_path: "/metrics_prometheus"
    static_configs:
      - targets: ["master-1:8051", "master-2:8051", "master-3:8051"]
        labels:
          group: "masters"
      - targets: ["tserver-1:8050", "tserver-2:8050", "tserver-3:8050"]
        labels:
          group: "tservers"

If your cluster uses automated service discovery, use the http_sd_configs parameter instead of static_configs.

The following example shows an HTTP service discovery configuration:


global:
  scrape_interval: 15s
  evaluation_interval: 15s

scrape_configs:
  - job_name: "kudu"
    metrics_path: "/metrics_prometheus"
    http_sd_configs:
      - url: "http://master-1:8051/prometheus-sd"
      - url: "http://master-2:8051/prometheus-sd"
      - url: "http://master-3:8051/prometheus-sd"

Start the Prometheus service.
If you use a custom configuration file path, run the following command:
```
prometheus --config.file="<custom_path>/prometheus.yml"
```
Access the Prometheus user interface and check the Status Targets page to verify that the Kudu daemons are connected.