kubectl常用命令

👈🏻 Select language

kubectl常用命令

    kubectl api-resources --namespaced=false
    kubectl api-resources --namespaced=true
    # -R表示递归目录下所有配置
    kubectl apply -R -f configs/


    kubectl get cs
    kubectl get svc --sort-by=.metadata.creationTimestamp
    # 查看节点
    kubectl get no --sort-by=.metadata.creationTimestamp
    kubectl drain <node-name>
    kubectl taint nodes node1 key=value:NoSchedule


    kubectl get ing pdd --n java
    # 不调度
    kubectl cluster-info dump


    kubectl get po --field-selector spec.nodeName=xxxx
    kubectl exec -it ng-57d74c8694-6cqnz sh  -n=java
    kubectl get pods --all-namespaces --field-selector spec.nodeName=<node> -o wide
    kubectl get pods -o wide --all-namespaces
    kubectl get po -l app=nginx -w
    kubectl delete po -l app=onekey-ali-web -n=$(namespace)
    kubectl get po --all-namespaces
    # 查看异常pod
    kubectl get po --all-namespaces --field-selector 'status.phase!=Running'

集群管理相关命令

应用管理相关

    kubectl top pod
    kubectl delete deployment,services -l app=nginx 
    kubectl scale deployment/nginx-deployment --replicas=2
    kubectl get svc --all-namespaces=true
    kubectl rollout undo deployment/nginx-deployment --to-revision=2
    kubectl autoscale deployment <deployment-name> --min=2 --max=5 --cpu-percent=80
    #来watch ReplicaSet的变化。
    kubectl get rs -w
    kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1
    # deployment "nginx-deployment" image updated
    kubectl set image deploy monitorapi-deployment  monitorapi=registry-vpc.cn-shenzhen.aliyuncs.com/amiba/monitorapi:1.2.4 -n=java

强制删除

有时删除pv/pvc时会有问题,这个使用得加2个命令参数--grace-period=0 --force

删除所有失败的pod

  kubectl get po --all-namespaces --field-selector 'status.phase==Failed'
  kubectl delete po  --field-selector 'status.phase==Failed'
  #模糊删除pod
  key=
  kgpo -n default | grep $key | awk '{print $1}' | xargs kubectl delete po -n1 -n default
  

事件

    kubectl get events  --field-selector involvedObject.kind=Service --sort-by='.metadata.creationTimestamp'
    kubectl get event --all-namespaces  --field-selector involvedObject.name=$po

参考链接:

kubernetes 节点维护 cordon, drain, uncordon

Recommended Tools

kubectx

kubectx: Used to switch cluster access

kubens: Used to switch default namespace

kubectl-aliases

kubectl command aliases

Auto-completion

zsh

source <(kubectl completion zsh)  # setup autocomplete in zsh into the current shell
echo "if [ $commands[kubectl] ]; then source <(kubectl completion zsh); fi" >> ~/.zshrc # add autocomplete permanently to your zsh shell

For other methods, see kubectl Cheat Sheet

Common kubectl Commands

    kubectl api-resources --namespaced=false
    kubectl api-resources --namespaced=true
    # -R means recursively apply all configs in the directory
    kubectl apply -R -f configs/


    kubectl get cs
    kubectl get svc --sort-by=.metadata.creationTimestamp
    # View nodes
    kubectl get no --sort-by=.metadata.creationTimestamp
    kubectl drain <node-name>
    kubectl taint nodes node1 key=value:NoSchedule


    kubectl get ing pdd --n java
    # Don't schedule
    kubectl cluster-info dump


    kubectl get po --field-selector spec.nodeName=xxxx
    kubectl exec -it ng-57d74c8694-6cqnz sh  -n=java
    kubectl get pods --all-namespaces --field-selector spec.nodeName=<node> -o wide
    kubectl get pods -o wide --all-namespaces
    kubectl get po -l app=nginx -w
    kubectl delete po -l app=onekey-ali-web -n=$(namespace)
    kubectl get po --all-namespaces
    # View abnormal pods
    kubectl get po --all-namespaces --field-selector 'status.phase!=Running'

    kubectl top pod
    kubectl delete deployment,services -l app=nginx 
    kubectl scale deployment/nginx-deployment --replicas=2
    kubectl get svc --all-namespaces=true
    kubectl rollout undo deployment/nginx-deployment --to-revision=2
    kubectl autoscale deployment <deployment-name> --min=2 --max=5 --cpu-percent=80
    # To watch ReplicaSet changes.
    kubectl get rs -w
    kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1
    # deployment "nginx-deployment" image updated
    kubectl set image deploy monitorapi-deployment  monitorapi=registry-vpc.cn-shenzhen.aliyuncs.com/amiba/monitorapi:1.2.4 -n=java

Force Delete

Sometimes there are problems when deleting pv/pvc. In this case, you need to add 2 command parameters --grace-period=0 --force

Delete All Failed Pods

  kubectl get po --all-namespaces --field-selector 'status.phase==Failed'
  kubectl delete po  --field-selector 'status.phase==Failed'
  # Fuzzy delete pods
  key=
  kgpo -n default | grep $key | awk '{print $1}' | xargs kubectl delete po -n1 -n default
  

Events

    kubectl get events  --field-selector involvedObject.kind=Service --sort-by='.metadata.creationTimestamp'
    kubectl get event --all-namespaces  --field-selector involvedObject.name=$po

Reference links:

kubernetes Node Maintenance cordon, drain, uncordon

推奨ツール

kubectx

kubectx：クラスターアクセスの切り替えに使用

kubens：デフォルトのnamespaceの切り替えに使用

kubectl-aliases

kubectlコマンドのエイリアス

自動補完

zsh

source <(kubectl completion zsh)  # setup autocomplete in zsh into the current shell
echo "if [ $commands[kubectl] ]; then source <(kubectl completion zsh); fi" >> ~/.zshrc # add autocomplete permanently to your zsh shell

他の方法については、kubectl Cheat Sheetを参照してください。

kubectlのよく使うコマンド

    kubectl api-resources --namespaced=false
    kubectl api-resources --namespaced=true
    # -Rはディレクトリ内のすべての設定を再帰的に適用することを意味します
    kubectl apply -R -f configs/


    kubectl get cs
    kubectl get svc --sort-by=.metadata.creationTimestamp
    # ノードを表示
    kubectl get no --sort-by=.metadata.creationTimestamp
    kubectl drain <node-name>
    kubectl taint nodes node1 key=value:NoSchedule


    kubectl get ing pdd --n java
    # スケジュールしない
    kubectl cluster-info dump


    kubectl get po --field-selector spec.nodeName=xxxx
    kubectl exec -it ng-57d74c8694-6cqnz sh  -n=java
    kubectl get pods --all-namespaces --field-selector spec.nodeName=<node> -o wide
    kubectl get pods -o wide --all-namespaces
    kubectl get po -l app=nginx -w
    kubectl delete po -l app=onekey-ali-web -n=$(namespace)
    kubectl get po --all-namespaces
    # 異常なpodを表示
    kubectl get po --all-namespaces --field-selector 'status.phase!=Running'

クラスター管理関連のコマンド

アプリケーション管理関連

    kubectl top pod
    kubectl delete deployment,services -l app=nginx 
    kubectl scale deployment/nginx-deployment --replicas=2
    kubectl get svc --all-namespaces=true
    kubectl rollout undo deployment/nginx-deployment --to-revision=2
    kubectl autoscale deployment <deployment-name> --min=2 --max=5 --cpu-percent=80
    # ReplicaSetの変化を監視するために。
    kubectl get rs -w
    kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1
    # deployment "nginx-deployment" image updated
    kubectl set image deploy monitorapi-deployment  monitorapi=registry-vpc.cn-shenzhen.aliyuncs.com/amiba/monitorapi:1.2.4 -n=java

強制削除

pv/pvcを削除する際に問題が発生することがあります。この場合、2つのコマンドパラメータ--grace-period=0 --force を追加する必要があります。

すべての失敗したPodを削除

  kubectl get po --all-namespaces --field-selector 'status.phase==Failed'
  kubectl delete po  --field-selector 'status.phase==Failed'
  # あいまいな削除pod
  key=
  kgpo -n default | grep $key | awk '{print $1}' | xargs kubectl delete po -n1 -n default
  

イベント

    kubectl get events  --field-selector involvedObject.kind=Service --sort-by='.metadata.creationTimestamp'
    kubectl get event --all-namespaces  --field-selector involvedObject.name=$po

参考リンク：

kubernetes ノードメンテナンス cordon, drain, uncordon

Общие команды kubectl

    kubectl api-resources --namespaced=false
    kubectl api-resources --namespaced=true
    # -R означает рекурсивно применить все конфигурации в каталоге
    kubectl apply -R -f configs/


    kubectl get cs
    kubectl get svc --sort-by=.metadata.creationTimestamp
    # Просмотр узлов
    kubectl get no --sort-by=.metadata.creationTimestamp
    kubectl drain <node-name>
    kubectl taint nodes node1 key=value:NoSchedule


    kubectl get ing pdd --n java
    # Не планировать
    kubectl cluster-info dump


    kubectl get po --field-selector spec.nodeName=xxxx
    kubectl exec -it ng-57d74c8694-6cqnz sh  -n=java
    kubectl get pods --all-namespaces --field-selector spec.nodeName=<node> -o wide
    kubectl get pods -o wide --all-namespaces
    kubectl get po -l app=nginx -w
    kubectl delete po -l app=onekey-ali-web -n=$(namespace)
    kubectl get po --all-namespaces
    # Просмотр аномальных подов
    kubectl get po --all-namespaces --field-selector 'status.phase!=Running'

Команды, связанные с управлением кластером

Связанные с управлением приложениями

    kubectl top pod
    kubectl delete deployment,services -l app=nginx 
    kubectl scale deployment/nginx-deployment --replicas=2
    kubectl get svc --all-namespaces=true
    kubectl rollout undo deployment/nginx-deployment --to-revision=2
    kubectl autoscale deployment <deployment-name> --min=2 --max=5 --cpu-percent=80
    # Для наблюдения за изменениями ReplicaSet.
    kubectl get rs -w
    kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1
    # deployment "nginx-deployment" image updated
    kubectl set image deploy monitorapi-deployment  monitorapi=registry-vpc.cn-shenzhen.aliyuncs.com/amiba/monitorapi:1.2.4 -n=java

Принудительное удаление

Иногда возникают проблемы при удалении pv/pvc. В этом случае нужно добавить 2 параметра команды --grace-period=0 --force

Удалить все неудачные поды

  kubectl get po --all-namespaces --field-selector 'status.phase==Failed'
  kubectl delete po  --field-selector 'status.phase==Failed'
  # Нечеткое удаление подов
  key=
  kgpo -n default | grep $key | awk '{print $1}' | xargs kubectl delete po -n1 -n default
  

События

    kubectl get events  --field-selector involvedObject.kind=Service --sort-by='.metadata.creationTimestamp'
    kubectl get event --all-namespaces  --field-selector involvedObject.name=$po

Ссылки:

Обслуживание узлов kubernetes cordon, drain, uncordon

推荐工具

自动完成