Kubernetes

k8s

kubectl

api-resources

kubectl api-resources List available resources

explain

kubectl explain <pods> --recursive Explain something

get

kubectl get <all/pods/replicationcontroller/replicaset/deployment/ns/svc/nodes Show all/replicationcontroller/replicaset/deployment/namespace/service objects
kubectl get pods -o wide Show all pods with more info

run/create

kubectl run <pod-name> --image=nginx Run new nginx pod
kubectl create -f <defintion.yaml> Create new object from <definition.yaml> file
kubectl run <pod-name> --image=nginx --dry-run=client -o yaml > sample.yaml Write a defintion file without starting the pod

describe/edit/delete

kubectl describe pod <pod-name> Describe pod
kubectl edit pod <pod-name> Edit pod with editor
kubectl delete pod <pod-name> Delete pod from default namespace

scale

kubectl scale --replicas=5 replicaset <replicaset-name> Edit the current replicas without editing the file (fleeting?).

explain

kubectl explain <something> Explain <something>

label

kubectl label <something> Edit labels <something>

Kubernetes Definition Files

“Can I have more than one of these?”
Yes → use -
No → don’t use -

Example .yaml for Pod

This configuration defines a pod named "myapp-pod" that runs a single container using the nginx image.

apiVersion: v1
kind: Pod
metadata:
  name: myapp-pod
  labels:
    app: myapp
    type: front-end
spec:
  tolerations:
  - key: "spray"
    operator: "Equal"
    value: "mortein"
    effect: "NoSchedule"
  containers:
  - name: nginx-container
    image: nginx

Example .yaml for Replication Controller

This configuration defines a replication controller for the example "myapp-pod" nginx pod.

apiVersion: v1
kind: ReplicationController
metadata:
  name: myapp-rc
  labels:
      app: myapp
      type: front-end
spec:
  template:
    metadata:
      name: myapp-pod
      labels:
        app: myapp
        type: front-end
    spec:
      containers:
      - name: nginx-container
        image: nginx
  replicas: 3

Example .yaml for ReplicaSet

This configuration defines a replica set for the example "myapp-pod" nginx pod.

apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: myapp-replicaset
  labels:
      app: myapp
      type: front-end
spec:
  template:
    metadata:
      name: myapp-pod
      labels:
        app: myapp
        type: front-end
    spec:
      containers:
      - name: nginx-container
        image: nginx
  replicas: 3
  selector:
    matchLabels:
      type: front-end

Example .yaml for Deployment

This configuration defines a deployment for the example "myapp-pod" nginx pod.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-replicaset
  labels:
      app: myapp
      type: front-end
spec:
  template:
    metadata:
      name: myapp-pod
      labels:
        app: myapp
        type: front-end
    spec:
      containers:
      - name: nginx-container
        image: nginx
  replicas: 3
  selector:
    matchLabels:
      type: front-end

Example .yaml for Namespace

This configuration defines a new namespace dev.

apiVersion: v1
kind: Namespace
metadata:
  name: dev

Command: kubectl create namespace dev

Example .yaml for Resource Quota

This configuration defines a resource quota compute-quota for namespace dev.

apiVersion: v1
kind: ResourceQuota
metadata:
  name: compute-quota
  namespace: dev
spec:
  hard:
    pods: "10"
    requests.cpu: "4"
    requests.memory: 5Gi
    limit.cpu: "10"
    limit.memory: 10Gi

Imperative Commands

Can be tested with: --dry-run=client
kubectl run --image=nginx nginx
kubectl create deployment --image=nginx nginx
kubectl expose deployment nginx --port 80
kubectl edit deployment nginx
kubectl scale deployment nginx --replicas=5
kubectl set image deployment nginx nginx=nginx:1.18

Taint

kubectl taint node node01 key=value:NoSchedule Add some taint to node01

Node Affinity

requiredDuringSchedulingIgnoredDuringExecution
preferredDuringSchedulingIgnoredDuringExecution

Deployment:
  spec:
    template:
      spec:
        affinity:
          nodeAffinity:
            requiredDuringSchedulingIgnoredDuringExecution:
              nodeSelectorTerms:
              - matchExpressions:
                - key: color
                  operator: In/NotIn/Exists
                  values:
                  - blue

Resource Requests & Limits

By default, a container has no limit to the resources it can consume on a node and may suffocate other pods or services on the same node.
Usual best practice: Set Requests with no limits.

spec:
  containers:
    resources:
      requests:
        cpu: 2
        memory: "4Gi"
       limits:
         cpu: 5
         memory: "8Gi"

LimitRange (CPU), applies to namespaces:

apVersion: v1
kind: LimitRange
metadata:
  name: cpu-resource-constraint
spec:
  limits:
  - default: (Limit)
      cpu: 500m
    defaultRequest: (Request)
      cpu: 500m
    max: (Limit)
      cpu: "1"
    min: (Request)
      cpu: 100m
    type:
      Container

LimitRange (CPU), applies to namespaces:

apVersion: v1
kind: LimitRange
metadata:
  name: memory-resource-constraint
spec:
  limits:
  - default: (Limit)
      memory: 1Gi
    defaultRequest: (Request)
      memory: 1Gi
    max: (Limit)
      memory: 1Gi
    min: (Request)
      memory: 500Mi
    type:
      Container

ResourceQuota, applies to namespaces:

apVersion: v1
kind: ResourceQuota
metadata:
  name: my-resource-quota
spec:
  hard:
    requests.cpu: 4
    requests.memory: 4Gi
    limits.cpu: 10
    limits.memory: 10Gi

DaemonSets

For monitoring or logs (kubeproxy). Creation is similar to ReplicaSet.

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: monitoring-daemon
spec:
  template:
    metadata:
      labels:
        app: monitoring-agent
    spec:
      containers:
      - name: monitoring-agent
        image: monitoring-agent
  selector:
    matchLabels:
      app: monitoring-agent

Priority Classes

System range: 2'147'483'648 - 2'000'000'000
Usable range: 1'000'000'000 - -2'147'483'648

apiVersion: scheduling.k8s.io/v1
kind: PriorityClass
metadata:
  name: high-priority
value: 1000000000
description: "Priority class for mission critical pods" (optional)
globalDefault: true (optional)
preemptionPolicy: PreemtLowerPriority (default)/ never (optional)

Assign by setting priorityClassName in pod definition under spec.

Admission Controllers

ps -ef | grep admission Show info about admission plugins

Performance and Monitoring

kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml
serviceaccount/metrics-server created

kubectl top <pod/node>

Logs

kubectl logs <pod/node>