🚀 Brimble Mini PaaS

A lightweight Platform-as-a-Service (PaaS) that enables application deployment, routing, and log streaming.

This project focuses on core deployment mechanics with minimal infrastructure, making it ideal for demonstrating backend architecture and system design fundamentals.

---

✨ Features

• ⚡ Deploy applications via API (file upload or git URL)
• 🌐 Dynamic routing using Caddy
• 📦 Docker-based container deployments
• 📡 Real-time log streaming (Server-Sent Events)
• 🔁 Lightweight async processing using queueMicrotask
• 🧠 Simple deployment orchestration
• ⚙️ Zero configuration required

---

🧱 Tech Stack

• Node.js (TypeScript)
• Express
• Docker
• Caddy
• Multer (file uploads)

---

🧠 Design Philosophy

This project intentionally avoids heavy infrastructure (like Redis or job queues) to keep the system:

• Simple
• Fast to run locally
• Focused on core PaaS concepts

Async work is handled using the native event loop:

queueMicrotask(() => {
  void this.deploymentManager.runDeployment(deployment.id, request);
});

This ensures non-blocking execution without introducing external dependencies.

---

🏗️ Architecture Overview

Client → API → Deployment Manager → Docker → Caddy → Deployed App

Logs → Server-Sent Events (SSE) → Client

---

🚀 Getting Started

Prerequisites

• Node.js
• Docker

---

Installation

git clone https://github.com/Simplecodez/brimble-mini-paas.git
cd brimble-mini-paas
sudo docker compose up -d

then open http://localhost:8081 to test the app

No environment variables required — defaults are preconfigured.

---

📡 API Endpoints

Deployments

Method	Endpoint	Description
POST	/api/deployments	Upload (git url) & deploy a project
GET	/api/deployments	List all deployments
GET	/api/deployments/:id	Get a single deployment

📦 Upload Format

• Content-Type: multipart/form-data
• Field name: projectArchive
• Expected: zipped project directory

---

Logs (Real-time)

Method	Endpoint	Description
GET	/api/events	Stream deployment logs (SSE)

This endpoint uses Server-Sent Events (SSE) to stream logs in real time.

---

📦 Deployment Flow

1. Client uploads project archive or git url
2. Deployment record is created
3. Deployment runs asynchronously via queueMicrotask
4. Docker image is built
5. Container is started
6. Caddy dynamically routes traffic
7. Logs are streamed via SSE

---

🧩 Routing Structure

// Deployments
POST   /api/deployments
GET    /api/deployments
GET    /api/deployments/:id

// Logs (SSE)
GET    /api/events

---

⚠️ Limitations (Intentional for Assessment)

• No distributed queue system
• No authentication
• No autoscaling

These trade-offs were made to prioritize clarity and core functionality.

---

🚀 Future Improvements

• Introduce job queue (BullMQ / Redis or other job queues)
• Rollbacks
• Custom domains & SSL
• Horizontal scaling

---

📄 License

MIT

---

👨‍💻 Author

Built by Simplecodez