How cos-in/iluvatar-faas works
Main Components and Architecture
The system should be represented with these key components:
a) Core Components:
• Controller (iluvatar_controller)
• Worker Nodes (iluvatar_worker)
• Load Generator (iluvatar_load_gen)
• Energy Monitor (iluvatar_energy_mon)
b) Supporting Services:
• Container Runtime (containerd/docker)
• Network Management
• Resource Management (CPU/GPU/Memory)
• Monitoring & Metrics (InfluxDB)
Component Relationships
a) Control Flow:
• Controller → Workers (load balancing, registration)
• Load Generator → Controller (workload distribution)
• Workers → Container Runtime (function execution)
• Energy Monitor → Workers (resource monitoring)
b) Data Flow:
• Function invocation path
• Metrics collection path
• Resource allocation flow
Detailed Component Breakdown
For each major component, include:
Worker Node:
• Container Pool
• Invocation Queue
• Resource Managers (CPU/GPU)
• Network Namespace Manager
• Health Monitor
Controller:
• Load Balancer
• Registration Service
• Health Check Service
• Async Invoke Handler
Implementation Details to Highlight
Key architectural features:
• Support for both simulation and live execution
• Load generation capabilities
• Energy monitoring integration
• Container runtime abstraction
• Queue management systems
• Resource isolation mechanisms
Load balancing and scaling mechanisms
This will create a comprehensive yet clear representation of Ilúvatar's architecture that highlights its research-focused design while showing all essential components and their interactions.