k8s 常用调式方法

• 本地数据库连接：

  mysql -h127.0.0.1 -P 3306 -uroot -pcorona@2020@xxxxx

• 常用 k8s kubectl 命令

  kubectl get po # 查看目前所有的pod
  kubectl get rs # 查看目前所有的replica set
  
  kubectl describe po my-nginx # 查看my-nginx pod的详细状态
  kubectl describe rs my-nginx # 查看my-nginx replica set的详细状态
  kubectl describe deployment my-nginx # 查看my-nginx deployment的详细状态
  kubectl get eventskubectl get events查看相关事件
  

• 查看任务：

  kubectl get deployment

• 删除任务

  kubectl delete deployment my-nginx

• 查看集群

  kubectl get pods -o wide -n kube-system ## kube-system 为用户空间"namespace"

• 查看 pod 环境变量：

  kubectl exec -it 35abba0b-6f16-4336-9d2e-a9a5326c209e -- env

• 设置kubectl 快捷键

  Vim ~/.bashrc
  
  alias kdpk='kubectl describe pod -n kube-system'
  alias kdpd='kubectl describe pod -n default'
  alias kdpc='kubectl describe pod -n custom-metrics'
  
  alias kdn='kubectl describe node'
  alias kddsk='kubectl describe ds -n kube-system'
  alias kddpk='kubectl describe deployment -n kube-system'
  
  alias kld='kubectl logs -n default --tail 20'
  alias klc='kubectl logs -n custom-metrics --tail 20'
  alias klk='kubectl logs -n kube-system --tail 20'
  
  alias kek='kubectl exec -n kube-system -it' 
  alias ked='kubectl exec -n default -it'  
  alias kec='kubectl exec -n custom-metrics -it' 
  
  alias kgda='kubectl get deployment --all-namespaces'
  alias kgpa='kubectl get po --all-namespaces'
  alias kgpa2='kubectl get pod -o=custom-columns=Namespace:.metadata.namespace,NAME:.metadata.name,STATUS:.status.phase,NODE:.spec.nodeName --all-namespaces'
  alias kgca='kubectl get configmap --all-namespaces'
  alias kgn='kubectl get nodes'
  alias kgdsa='kubectl get ds --all-namespaces'
  
  alias kgpyd='kubectl get pod -n default -o yaml'
  alias kgpyk='kubectl get pod -n kube-system -o yaml'
  
  alias kdtpd='kubectl delete pod -n default'
  alias kdtpk='kubectl delete pod -n kube-system'
  alias kdtdsk='kubectl delete ds -n kube-system'
  
  Source ~/.bashrc
  

• 指定容器资源占用：

  docker run -it --rm ivotron/stress-ng:v0.07.29-arm64v8 /usr/bin/stress-ng --cpu 2 --cpu-load 100 --metrics-brie

• 查看 pods

  kgpa | grep mysql

• 查看pods 详情：

  kdpk mysql-xwmtw

• 查看 pod 错误log ：

  kld addon-custom-user-dasboard-backend-vvfds

• 确认服务器中是否有nvidia显卡的命令

  lspci -vnnn | perl -lne 'print if /^\d+\:.+(\[\S+\:\S+\])/' | grep VGA | grep -i NVIDIA

• 等待kubectl执行：wait

  1. Wait for the pod “busybox1” to contain the status condition of type “Ready”.

     kubectl wait --for=condition=Ready pod/grafana

  2. Wait for the pod “busybox1” to be deleted, with a timeout of 60s, after having issued the “delete” command.

     kubectl delete pod/busybox1
     kubectl wait --for=delete pod/busybox1 --timeout=60s
     

  • 示例，重启监控组件如grafana,或openstry,如果立即stop/start，pod的服务会因主机端口占用启用失败；此时可以增加wait选项等待stop操作完成

    kubectl wait --for=delete pod/openstry* --timeout=60s -n kube-system

2. 查看并释放端口

   netstat -tln | grep 8080
   lsof -i:8060
   

参考：
kubectl_wait