⚡ Infra-controller

The Central Nervous System of the FaaS Platform

Traffic Orchestration • Resource Governance • Real-time Monitoring

---

📖 Introduction

The Infra Controller acts as the high-throughput brain of the serverless architecture. Leveraging the non-blocking I/O of Node.js, it handles request authentication, rate limiting, and intelligent job distribution to the worker fleet.

It now supports Dynamic Resource Resizing, allowing users and Auto-Tuners to adjust function memory configuration on the fly without service interruption.

---

🏗️ Architecture

graph LR
    User[Client] -- "x-api-key" --> ALB[Load Balancer]
    ALB -- Port 8080 --> Controller[Node.js Controller]
    
    subgraph "Control Plane (Infra)"
    Controller -- "1. Upload (Zip)" --> S3[AWS S3]
    Controller -- "2. Metadata" --> DDB[DynamoDB]
    Controller -- "3. Rate Limit" --> Redis[(Redis Cluster)]
    Controller -- "4. Enqueue Job" --> SQS[AWS SQS]
    
    Controller -.-> Prom[Prometheus]
    Controller -.-> AINode[AI Node / Ollama]
    end

    SQS --> Worker[Worker Nodes]
    Worker -- "5. Result Pub" --> Redis
    Redis -- "6. Sub/Result" --> Controller
    Controller --> User

Loading

---

✨ Key Features

🛡️ Enterprise Security & Governance

• API Guard: Strict authentication via x-api-key header verification.
• Traffic Control: Atomic rate limiting using Redis Lua Scripts (e.g., 100 req/min depending on IP).
• Validation: Strict schema checks for input payloads and configuration updates.

🏥 Operational Stability (Resilience)

• Fail-Fast Startup: Immediately exits if critical environment variables (AWS credentials, DB Table) are missing to prevent "Zombie" states.
• Graceful Shutdown: Safely closes pending requests and DB connections upon SIGTERM (Zero Downtime deployment).
• Deep Health Check: /health endpoint validates Redis connectivity and Queue depth, not just process uptime.

🔭 Deep Observability

• RED Metrics: Exposes Requests, Errors, and Duration metrics via /metrics (Prometheus).
• Audit Logs: Archives full execution history to DynamoDB (InfraExecutionLogs) with auto-expiry (TTL).
• Tracing: End-to-end traceability with UUID-based Request IDs.

⚡ Dynamic Management (New)

• Hot-Swap Config: Update function memory (memoryMb) and description instantly via API.
• Auto-Tuner Support: Accepts optimization commands from the Auto-Tuner engine to resize memory with a single click.

---

🚀 Getting Started

Environment Setup (.env)

Variable	Description	Example
PORT	Service Port	8080
TABLE_NAME	Metadata Table	InfraFunctions
LOGS_TABLE_NAME	Logs Table	InfraExecutionLogs
SQS_URL	Task Queue URL	https://sqs...
REDIS_HOST	Redis Endpoint	localhost
INFRA_API_KEY	Master Auth Key	secret-key

Launch Commands

# 1. Install
npm install

# 2. Run (Dev)
node controller.js

# 3. Production (PM2)
pm2 start controller.js --name "faas-controller" -i max

---

📡 API Reference

Auth Header: x-api-key: <YOUR_SECRET_KEY>

1. Execute Function

POST /run Orchestrates a synchronous or asynchronous execution.

// Request
{ 
  "functionId": "func-123", 
  "inputData": { "msg": "hello" } 
}

// Response
{
  "status": "SUCCESS",
  "data": "Result...",
  "durationMs": 120
}

2. Update Configuration (New)

PUT /functions/:id Now supports On-the-fly Memory Resizing.

// Request
{
  "description": "Optimized by Auto-Tuner",
  "memoryMb": 128  // Updated from 512
}

3. Get Logs

GET /api/functions/:id/logs Fetch historical execution logs with pagination.

---

💻 Terminal Output Demo

Server Startup:

{"level":"INFO","msg":"Global Redis Connected Successfully"}
{"level":"INFO","msg":"Infra Controller v2.5 Started","port":8080}

Function Execution:

{"level":"INFO","msg":"Run Request","requestId":"a1b2...","functionId":"func-123","mode":"SYNC"}
{"level":"INFO","msg":"Execution Success","duration":120}

---

_{Built for Scalability and Control}