Blogshog

Overview

A community blogging platform built on the MERN stack. Write, read, and comment — with a clean API, flexible data model, and JWT auth under the hood.

This was my first real full-stack build. The goal was straightforward: a place to write and read blogs. But getting the data model right, wiring auth securely, and designing a scalable API surface turned it into a genuine learning ground.

System Design & Architecture

Blogshog follows a classic client-server architecture using the MERN stack:

Client (React)

• Component-driven UI for blogs, comments, and profiles
• Global state management using Redux
• Handles user interactions and API communication

Backend (Node.js + Express)

• RESTful API for:
  • Blog CRUD operations
  • Comment management
  • User authentication
• Middleware-based architecture for request handling and validation

Data Layer (MongoDB)

• Document-based schema designed for:
  • Blogs (content, author, metadata)
  • Comments (linked to blogs and users)
  • Likes (linked to blogs and users)
  • Users (profiles and authentication data)

---

Key Engineering Decisions

1. Document-Based Data Modeling

Decision: Use MongoDB with reference-based relationships.

Why:

• Embedding comments inside blog documents would cause them to grow unbounded
• Flexibility in handling varying blog content structures
• Using reference IDs (blogId, userId) keeps documents lean and queries predictable
• Simpler schema evolution compared to relational databases

2. RESTful API Design

Decision: Structure backend as modular REST endpoints.

Why:

• Clear separation of concerns
• Easier to extend with new features (bookmarks etc.)

3. Centralized State Management

Decision: Use Redux for managing authentication and blog state.

Why:

• Multiple components depend on shared data (user session, blog list, comments)
• Centralizing it in Redux made updates predictable and debugging straightforward.

4. JWT-Based Authentication

Decision: Use token-based authentication for securing APIs.

Why:

• Stateless and scalable
• Works well with REST APIs

Implementation:

• Issued JWTs on login/signup
• Protected routes via middleware validation

---

Challenges

Managing Relational Data in a NoSQL Database

MongoDB does not provide native joins like SQL systems.

Solution:

• Used reference IDs to link users, blogs, and comments
• Structured queries to fetch related data efficiently

Designing a Flexible Comment System

Comments must be tied to both users and blog posts.

Solution:

• Created a schema linking comment → blogId → userId
• Ensured consistent data retrieval and rendering

Securing User Authentication

Handling user sessions securely is critical in content platforms.

Solution:

• Implemented JWT-based authentication
• Protected sensitive routes via middleware

---

Learnings

This project focused on building data-driven, API-centric systems:

• Managing global application state effectively
• Designing schemas for relational data in NoSQL environments
• REST API structure matters a lot as a codebase grows
• Auth is easy to get wrong — JWT + middleware is a solid, repeatable pattern