k8s新增jupyter服务

k8s新增服务

常用命令

• kubectl apply -f xxxxxx.yaml # 部署资源，顺序：namespace -> pvc -> deployment -> service
• kubectl create namespace jupyter # 创建namespace
• kubectl get namespaces # 查看ns
• kubectl get pods -n jupyter # 查看pods
• kubectl logs jupyterlab -n jupyter # 查看pod日志
• kubectl get svc jupyter -n jupyter # 查看 svc 和 nodeport
• kubectl get deployment -n jupyter # 查看 deployment
• kubectl get pods -A # 查看所有ns下的pod
• kubectl get all -n jupyter # 查看ns下的所有资源
• kubectl describe [deployment|svc] jupyter -n jupyter # 详细信息，可以查看事件
• kubectl edit deployment jupyter -n jupyter # 命令行编辑yaml，保存后立即生效
• kubectl apply -f jupyter-deployment.yaml # 应用更新后的yaml

新增namespace

apiVersion: v1
kind: Namespace
metadata:name: jupyter
新增pvc
apiVersion: v1
kind: PersistentVolumeClaim
metadata:name: jupyter-work-pvcnamespace: jupyter
spec:accessModes:- ReadWriteOnceresources:requests:storage: 10Gi

新增deployment

apiVersion: apps/v1
kind: Deployment
metadata:name: jupyternamespace: jupyter  # 假设你已经创建了 jupyter 这个 Namespace
spec:replicas: 1selector:matchLabels:app: jupytertemplate:metadata:labels:app: jupyterspec:containers:- name: jupyterimage: pyflink:20250109  # 安装了python notebook 的镜像securityContext:privileged: true  # 对应 docker-compose 的 privileged: trueports:- containerPort: 8888  # Jupyter Notebook 端口volumeMounts:- name: work-volumemountPath: /work     # 挂载 /work 目录command: ["/bin/sh", "-c"]args:- |cd /nohup jupyter notebook --ip 0.0.0.0 --port 8888 --allow-root --NotebookApp.password=sha1:6587feaef3b1:6b243404e4cfaafe611fdf494ee71fdaa8c4a563 2>&1 > /work/.jupyter.log &# 保持主进程运行tail -f /dev/nullresources:requests:cpu: "2000m"memory: "4Gi"limits:cpu: "4000m"memory: "8Gi"volumes:- name: work-volumepersistentVolumeClaim:claimName: jupyter-work-pvc  # 挂载 PVC（持久化 /work 目录）

新增service

apiVersion: v1
kind: Service
metadata:name: jupyternamespace: jupyter
spec:selector:app: jupyterports:- name: jupyter-portprotocol: TCPport: 8888targetPort: 8888type: NodePort  # 使用 NodePort 类型暴露服务