Kubernetes cluster monitoring (via Prometheus)

Dashboard

Monitors Kubernetes cluster using Prometheus. Shows overall cluster CPU / Memory / Filesystem usage as well as individual pod, containers, systemd services statistics. Uses cAdvisor metrics only.
Last updated: 5 months ago

Downloads: 1351

  • by_nodes.png
    by_nodes.png
  • systemd.png
    systemd.png
  • containers.png
    containers.png
  • pods.png
    pods.png
  • total.png
    total.png

release license k8s prometheus grafana link

Initial idea was taken from this dashboard and improved to exclude node-exporter dependency and to give more information about cluster state.

Requirements

You only need to have running Kubernetes cluster with deployed Prometheus. Prometheus will use metrics provided by cAdvisor via kubelet service (runs on each node of Kubernetes cluster by default) and via kube-apiserver service only.

Your Prometheus configuration has to contain following scrape_configs:

scrape_configs:
  - job_name: kubernetes-nodes-cadvisor
    scrape_interval: 10s
    scrape_timeout: 10s
    scheme: https  # remove if you want to scrape metrics on insecure port
    tls_config:
      ca_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
    bearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/token
    kubernetes_sd_configs:
      - role: node
    relabel_configs:
      - action: labelmap
        regex: __meta_kubernetes_node_label_(.+)
    metric_relabel_configs:
      - action: replace
        source_labels: [id]
        regex: '^/machine\.slice/machine-rkt\\x2d([^\\]+)\\.+/([^/]+)\.service$'
        target_label: rkt_container_name
        replacement: '${2}-${1}'
      - action: replace
        source_labels: [id]
        regex: '^/system\.slice/(.+)\.service$'
        target_label: systemd_service_name
        replacement: '${1}'

Features

  • Total and used cluster resources: CPU, memory, filesystem.
    And total cluster network I/O pressure.
    Total and used cluster resources
  • Kubernetes pods usage: CPU, memory, network I/O.
    Pods usage
  • Containers usage: CPU, memory, network I/O.
    Docker and rkt containers which runs on cluster nodes but outside Kubernetes are also monitored.
    Containers usage
  • systemd system services usage: CPU, memory.
    systemd usage
  • Showing all above metrics both for all cluster and each node separately.
    Filtering metrics by nodes

Troubleshooting

If filesystem usage panels display N/A, you should correct device=~"^/dev/[vs]da9$" filter parameter in metrics query with devices your system actually has.