Skip to main content
Version: 3.9

Pipeline Monitoring with Prometheus and Grafana

This documentation describes how to integrate tekton-pipelines metrics with Prometheus and Grafana monitoring stack.

Prerequisites​

Ensure the following requirements are met first before moving ahead:

Integration via Service Monitor​

To implement Pipeline Monitoring using the Service Monitor resource, follow the steps below:

  1. Create a Service Monitor resource file with the following configuration:

    apiVersion: monitoring.coreos.com/v1
    kind: ServiceMonitor
    metadata:
      labels:
        release: prom               # kube-prometheus-stack release name
      name: tekton-overview
    spec:
      endpoints:
        - interval: 10s
          path: /metrics
          port: http-metrics
          scrapeTimeout: 10s
      namespaceSelector:
        matchNames:
          - tekton-pipelines
      selector:
        matchLabels:
          app.kubernetes.io/component: controller
          app.kubernetes.io/part-of: tekton-pipelines

  2. Apply the created resource to the tekton-pipelines namespace:

    kubectl apply -f service-monitor.yaml -n tekton-pipelines

  3. Download the KubeRocketCi dashboard:

    a. Click on the dashboard menu;

    b. In the upper right corner, click on the New button. In the dropdown menu, click the Import;

    Import dashboard grafana

    c. Select the downloaded 18321_rev*.json file;

    Import dashboard grafana options

    d. Type the name of the dashboard;

    e. Select the folder for the dashboard;

    f. Change the UID if necessary;

    g. Click the Import button.

As soon as the dashboard procedure is completed, you can track the newcoming metrics in the dashboard menu:

Tekton dashboard