How to Deploy and Use Kubernetes Metrics Server for Cluster Monitoring

Comparison of Monitoring Tools

node-exporter collects server‑level metrics such as load average, filesystem usage, and memory information, similar to traditional host‑monitoring tools like Zabbix agent.

metric‑server (or heapster) retrieves CPU and memory usage metrics from the API server and forwards them to a storage backend (e.g., InfluxDB or a cloud provider), primarily to supply decision metrics for components like HPA.

kube‑state‑metrics focuses on exposing the current state of Kubernetes resources such as Deployments or DaemonSets.

Typical kube‑state‑metrics queries

How many Replicas are scheduled and how many are currently available?

How many Pods are in running, stopped, or terminated states?

How many times have Pods restarted?

How many Jobs are currently running?

These metrics are provided by kube‑state‑metrics. They are not included in metric‑server because the two components have fundamentally different goals.

metric‑server only fetches, formats, and writes existing data to a specific storage, essentially acting as a monitoring system.

kube‑state‑metrics snapshots Kubernetes status in memory and offers new metrics, but it cannot export them.

Deploying metric‑server

Download the deployment manifest:

wget https://github.com/kubernetes-sigs/metrics-server/releases/download/v0.3.7/components.yaml

Pull the metric‑server image:

# docker pull zhaoqinchang/metrics-server:0.3.7
0.3.7: Pulling from zhaoqinchang/metrics-server
9ff2acc3204b: Pull complete
9d14b55ff9a0: Pull complete
Digest: sha256:c0efe772bb9e5c289db6cc4bc2002c268507d0226f2a3815f7213e00261c38e9
Status: Downloaded newer image for zhaoqinchang/metrics-server:0.3.7

Edit components.yaml as shown (the essential parts are reproduced below):

# cat components.yaml
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: system:aggregated-metrics-reader
  labels:
    rbac.authorization.k8s.io/aggregate-to-view: "true"
    rbac.authorization.k8s.io/aggregate-to-edit: "true"
    rbac.authorization.k8s.io/aggregate-to-admin: "true"
rules:
- apiGroups: ["metrics.k8s.io"]
  resources: ["pods","nodes"]
  verbs: ["get","list","watch"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: metrics-server:system:auth-delegator
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: system:auth-delegator
subjects:
- kind: ServiceAccount
  name: metrics-server
  namespace: kube-system
---
# (additional RoleBinding, APIService, ServiceAccount, Deployment, Service, and ClusterRole definitions continue…)

Apply the manifest:

# kubectl apply -f components.yaml
clusterrole.rbac.authorization.k8s.io/system:aggregated-metrics-reader created
clusterrolebinding.rbac.authorization.k8s.io/metrics-server:system:auth-delegator created
rolebinding.rbac.authorization.k8s.io/metrics-server-auth-reader created
apiservice.apiregistration.k8s.io/v1beta1.metrics.k8s.io created
serviceaccount/metrics-server created
deployment.apps/metrics-server created
service/metrics-server created
clusterrole.rbac.authorization.k8s.io/system:metrics-server created
clusterrolebinding.rbac.authorization.k8s.io/system:metrics-server created

Verify that the metrics.k8s.io API group is registered:

# kubectl api-versions | grep metrics
metrics.k8s.io/v1beta1

Using kubectl top

The kubectl top command displays resource usage for Nodes and Pods, relying on the cluster’s metrics API.

Example – list node resource usage:

# kubectl top nodes
NAME      CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%
master    282m         14%    1902Mi           51%
node-02   70m          3%    1371Mi           37%
node-03   121m         1%    892Mi            11%

Example – list pods in the kube-system namespace:

# kubectl top pods -n kube-system
NAME                                 CPU(cores)   MEMORY(bytes)
etcd-master                          32m          300Mi
kube-apiserver-master                86m          342Mi
kube-controller-manager-master       30m          48Mi
kube-flannel-ds-l5ghn                5m           10Mi
kube-flannel-ds-rqlm2                4m           12Mi
kube-flannel-ds-v92r9                4m           14Mi
kube-proxy-7vjcv                     18m          15Mi
kube-proxy-xrz8f                     13m          21Mi
kube-proxy-zpwn6                     1m           14Mi
kube-scheduler-master                7m           17Mi
metrics-server-5549c7694f-7vb66      2m           14Mi

The kubectl top command provides a quick, built‑in way to monitor CPU and memory consumption of nodes and pods, making it essential for day‑to‑day Kubernetes operations.