📰 Scalable Article Hub

A scalable backend system for managing and serving large volumes of articles.
This project aims to explore high-performance backend architecture using Spring Boot, JPA, and JWT-based authentication.

---

🚀 Project Status – Week 1 ✅

• ✅ User & Post entity design (JPA)
• ✅ Basic CRUD for User and Post
• ✅ JWT-based authentication implemented
• ✅ Swagger (OpenAPI) documentation integrated
• ✅ Postman API testing completed
• ✅ @Transactional issue resolved and documented

🚀 Project Status – Week 2 ✅

• ✅ Large Dataset Generation & Insertion

  • Generated 10,000+ articles using Faker and inserted them into MySQL
  • Indexed all records into Elasticsearch for search performance testing

• ✅ Search API Implementation

  • MySQL Search: JPA with findByTitleContainingOrContentContaining + pagination & sorting
  • Elasticsearch Search: Java API Client with MultiMatchQuery (title, content) + author/date filters

• ✅ Performance Testing with Postman

  • Designed 5 test scenarios changing keyword, page, and size parameters
  • Measured response time for both MySQL and Elasticsearch

• ✅ Performance Comparison Results

  Request No.	MySQL Time (ms)	Elasticsearch Time (ms)
  1 (keyword=love)	193	37
  2 (keyword=kill)	186	31
  3 (keyword=story)	143	25
  4 (page=1)	104	25
  5 (size=5)	137	26
  Average	152.6	28.8

  ✅ Conclusion: Elasticsearch achieved ~5x faster average response times compared to MySQL.

🚀 Project Status – Week 3 ✅

• ✅ Redis Caching Integration

  • Redis container launched via docker-compose
  • Integrated Redis into Spring Boot for view count caching
  • Implemented scheduler to periodically flush cached deltas back to MySQL

• ✅ Cache Performance Testing (JMeter)

  • Designed tests with 100 concurrent users and different access patterns:
    • Hot (few popular articles repeatedly requested)
    • Random (uniform random requests)
    • Zipf (realistic skewed distribution)
    • Cold (no cache preloaded) vs Warm (cache preloaded)

• ✅ Cache Hit/Miss Results

  Scenario	Cold (Hits/Misses)	Warm (Hits/Misses)
  Hot	0 / 100	99 / 1
  Random	0 / 100	50 / 50
  Zipf	50 / 50	91 / 9

  ✅ Insights

  • Caching is most effective under Hot access pattern (~99% hit ratio).
  • Zipf distribution also benefits strongly from caching.
  • Random distribution shows limited caching gains, as expected.

• ✅ Visualization

  

  (Cache hit ratios for Hot, Random, Zipf under Cold vs Warm conditions)

---

🛠 Tech Stack

Layer	Tech Used
Backend	Spring Boot, Spring Security
Database	MySQL (JPA, Hibernate), Elasticsearch 8.x
Auth	JWT (access token only)
Docs/Test	Swagger (springdoc), Postman
Build Tool	Maven

---

📂 Current Structure

src/
└── main/
    ├── java/com/euni/articlehub/
    │   ├── controller/
    │   ├── service/
    │   ├── repository/
    │   ├── document/          # Elasticsearch documents
    │   ├── dto/
    │   ├── entity/
    │   ├── filter/
    │   └── util/
    └── resources/
        ├── application.yml
        └── static/templates (future)    

---

⚙️ How to Run

# Clone the project
git clone https://github.com/your-username/scalable-article-hub.git

# Navigate into the project
cd scalable-article-hub

# Run the Spring Boot application
./mvnw spring-boot:run

Swagger UI available at:

👉 http://localhost:8080/swagger-ui/index.html

Kibana (Elasticsearch monitoring) available at:

👉 http://localhost:5601

Redis (default port):

👉 redis://localhost:6379

RedisInsight (Redis GUI):

👉 http://localhost:5540

---

💼 About This Project

This project is part of my backend engineering portfolio, aimed at demonstrating:

• Scalable API development with Java & Spring Boot

• Performance optimization using Elasticsearch

• Secure architecture with JWT authentication

• Practical API testing using Postman