k8s常用命令笔记

一些自用的kubectl命令

重启pod

NAME_SPACE=anyonedev
kubectl get pod podname -n=${NAME_SPACE} -o yaml | kubectl replace --force -f  - 

强制删除pod

解决：加参数--force --grace-period=0
grace-period表示过渡存活期，默认30s，在删除POD之前允许POD慢慢终止其上的容器进程，从而优雅退出
0表示立即终止POD

kubectl delete pod <your-pod-name> -n=<name-space> --force --grace-period=0

根据状态过滤批量操作

过滤条件： ImagePullBackOff|CrashLoopBackOff|Evicted|Terminating

NAME_SPACE=anyonedev

## 删除所有ImagePullBackOff或CrashLoopBackOff的pod
kubectl get pods -n=${NAME_SPACE} | grep -E 'ImagePullBackOff|CrashLoopBackOff'  | awk '{print $1}' | xargs kubectl delete pod  -n=${NAME_SPACE}

## 重启所有evicted或Terminating的pod
kubectl get pods -n=${NAME_SPACE} | grep -E 'Evicted|Terminating'  | awk '{print $1}' | xargs kubectl get pod  -n=${NAME_SPACE} -o yaml | kubectl replace --force -f  - 

pod-forward

kubectl port-forward --address 0.0.0.0 pod/pod名称 暴露端口:内部端口
kubectl port-forward --address 0.0.0.0 service/service名称 暴露端口:内部端口

xargs查看pod日志

NAME_SPACE=bigdata
POD_NAME="kafka-clean"

kubectl -n=${NAME_SPACE} get pods | grep $POD_NAME | awk '{print $1}' | xargs  kubectl -n=${NAME_SPACE} logs

生成kubernetes集群最高权限admin用户的token

参考：https://jimmysong.io/kubernetes-handbook/guide/auth-with-kubeconfig-or-token.html

vim lin-admin-role.yaml

粘贴如下内容

kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1beta1
metadata:
  name: lin-admin
  annotations:
    rbac.authorization.kubernetes.io/autoupdate: "true"
roleRef:
  kind: ClusterRole
  name: cluster-admin
  apiGroup: rbac.authorization.k8s.io
subjects:
- kind: ServiceAccount
  name: lin-admin
  namespace: kube-system
---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: lin-admin
  namespace: kube-system
  labels:
    kubernetes.io/cluster-service: "true"
    addonmanager.kubernetes.io/mode: Reconcile

# 创建
kubectl create -f lin-admin-role.yaml
# 获取admin-token的secret名字
kubectl -n kube-system get secret|grep lin-admin-token
# 获取token的值
kubectl -n kube-system describe secret lin-admin-token-rhz5b

Q.E.D.