自架 JupyterHub 安裝 helm & 透過 helm 安裝 JupyterHub

我都是抄 Zero to JupyterHub with Kubernetes　ＸＤＤ

Setup k8s & helm

• k8s
  • docker desktop 解決
• helm
  • mac
    • brew install helm
  • windows
    • choco

      • install choco

            Set-ExecutionPolicy Bypass -Scope Process -Force; [System.Net.ServicePointManager]::SecurityProtocol = [System.Net.ServicePointManager]::SecurityProtocol -bor 3072; iex ((New-Object System.Net.WebClient).DownloadString('https://community.chocolatey.org/install.ps1'))
        

      • choco install kubernetes-helm

Set JupyterHub

看 jupyterhub helm chart version: JupyterHub’s Helm chart repository

• Initialize a Helm chart configuration file

  • *1. 讓 你電腦的 helm 知道你要安裝 JupyterHub Helm chart repository

       helm repo add jupyterhub https://jupyterhub.github.io/helm-chart/
       helm repo update
    

  • *2. 安裝 helm 的 chart install jupyterhub/jupyterhub

    • 官方這樣寫

          helm upgrade --cleanup-on-fail \
           --install <helm-release-name> jupyterhub/jupyterhub \
           --namespace <k8s-namespace> \
           --create-namespace \
           --version=<chart-version> \
           --values config.yaml
      

      • <helm-release-name> 你自己可以亂取 helm release name

      • <k8s-namespace> 你自己可以亂取 k8s namespace

      • <chart-version> Helm chart 版本,不是 jupyterHub 的唷!

      helm upgrade --cleanup-on-fail --install jub1 jupyterhub/jupyterhub --namespace jhub --create-namespace --version=1.2.0 --values config.yaml
    

    
      PS E:\yuting-project\ai-side-project\yuting-jupyter-hub> helm upgrade --cleanup-on-fail --install jub1 jupyterhub/jupyterhub --namespace jhub --create-namespace --version=1.2.0 --values config.yaml
      Release "jub1" does not exist. Installing it now.
      NAME: jub1
      LAST DEPLOYED: Fri Jan 14 09:53:25 2022
      NAMESPACE: jhub
      STATUS: deployed
      REVISION: 1
      TEST SUITE: None
      NOTES:
      Thank you for installing JupyterHub!
              
      Your release is named "jub1" and installed into the namespace "jhub".
              
      You can check whether the hub and proxy are ready by running:
              
       kubectl --namespace=jhub get pod
              
      and watching for both those pods to be in status 'Running'.
              
      You can find the public (load-balancer) IP of JupyterHub by running:
              
        kubectl -n jhub get svc proxy-public -o jsonpath='{.status.loadBalancer.ingress[].ip}'
              
      It might take a few minutes for it to appear!
              
      To get full information about the JupyterHub proxy service run:
              
        kubectl --namespace=jhub get svc proxy-public
              
      If you have questions, please:
              
        1. Read the guide at https://z2jh.jupyter.org
        2. Ask for help or chat to us on https://discourse.jupyter.org/
        3. If you find a bug please report it at https://github.com/jupyterhub/zero-to-jupyterhub-k8s/issues
    
    

  • *3 看有沒有成功跑起來

    • kubectl get service --namespace <<k8s-namespace>>

    

    • 看 proxy-public 那項 打 EXTERNAL-IP 那個項目 看看有愛情的登入畫面~~ 有的畫就成功拉

    

學習後記

1. 搞懂 整個 jupyterhub 在 k8s 裡面的機制和樣子
   • service
   • pod
   • 玩了一堆 要砍 pod
     • kubectl delete all --all -n {namespace}
2. 畫出來 :) 更有學習的感覺!
3. 好爽~~

Reference

• Installing JupyterHub¶