How manics/jupyterhub-virtualdesktopspawner works
Key Components to Include:
Core Infrastructure Layer
• Kubernetes Cluster (primary hosting platform)
• AWS EC2 (for Windows VMs)
• Networking components (VPC, Subnets)
Main Application Components
• JupyterHub (central orchestrator)
• Virtual Desktop Spawner Service
• Linux Desktop Spawner (Container-based)
• Windows Desktop Spawner (EC2-based)
• Apache Guacamole (remote desktop access service)
Authentication & Authorization
• AWS IAM/IRSA integration
• Kubernetes Service Accounts
• SSM (AWS Systems Manager for Windows instances)
Component Relationships
• User → Apache Guacamole (browser access)
• JupyterHub → Virtual Desktop Spawner
• Virtual Desktop Spawner → Two paths:
a. Linux path: Direct to Kubernetes Containers
b. Windows path: To AWS EC2 via AWS API
Specific Implementation Details
• Separation between the two spawner types (KubeEC2Spawner and container-based)
• IAM role and policy connections
• VPC/Subnet configuration for EC2 instances
• Guacamole connection broker
Additional Elements
• Version compatibility notes
• Development status (as it's under development)
• Configuration file references (z2jh-linuxdesktops.yaml, z2jh-windowsdesktops.yaml)
• The hybrid nature of the solution (container + VM based)
• Security boundaries and authentication flows
• Clear separation of Linux and Windows paths
• Integration points with cloud services
Remember to:
• Keep the main flow clear and easy to follow
• Necessary security components
• Scaling capabilities
• Failure recovery paths
• Development/beta components