An advanced, production-grade URL shortening service built for internet-scale systems. Designed with scalability, observability, and extensibility in mind — yet still under active development with new features being added regularly.
⚠️ Note: This project is still in development. More features and improvements are being added continuously. Stay tuned!
A URL shortener transforms long URLs into short, manageable links that are easier to share and track. For example:
Original: https://example.com/urls/www.google.com/about/ndksnadkdasd..
Shortened: https://yourshorturl.com/xYz1K8
When the short URL is accessed, the service redirects the user to the original long URL — possibly validating, tracking, and applying security or expiration rules in between.
This URL shortener is built with large-scale production usage in mind:
- 🔥 100M Daily Active Users
- 📈 Read:Write ratio of 100:1
- 🧠 Data retention: 5 years
- 🧲 1M write requests/day
- 📊 Optimized for heavy read traffic and long-term analytics
- 🔗 Long URL shortening
- ↪️ Short URL redirection
- 🏷️ Custom aliases for URLs
- 📷 QR code generation
- ⏳ URL expiration (time-based and click-based)
- ✅ URL validation and reachability checks
- 🔐 Password-protected short URLs
- 🍪 Cookies-based session management and Remember Me
- 📊 Click-based expiration
- 📣 Social media share tracking
- 🧠 Link versioning for A/B testing and rollback
- 📊 Analytics (via Kafka):
- Click counts
- Device, browser, OS and other info
- Geographic location
- 🌸 Bloom Filter for fast duplicate checks
- ✨ Redis caching for lightning-fast lookups
- 🧱 PostgreSQL as the persistent backend
- 🧪 Server-side rendering with HTML, Tailwind, JavaScript + Thymeleaf for user-facing pages (like password entry)
To ensure distributed-safe short codes, the system uses:
- ❄️ Snowflake IDs (timestamp + machine ID)
- Encoded via Base62 (A–Z, a–z, 0–9) for shorter, URL-friendly codes
- ❌ Avoided manually configured machine IDs via
application.ymlto reduce human error
-
Uses a multi-layered approach:
- Bloom Filter (existence check)
- Redis (hot read/write cache)
- PostgreSQL (source of truth)
⚠️ In future iterations, we can replace PostgreSQL with Cassandra or DynamoDB for true horizontal scalability.
| Layer | Tech |
|---|---|
| Backend | Java 21, Spring Boot |
| Async Messaging | Apache Kafka, Kafdrop |
| Database | PostgreSQL |
| Cache | Redis |
| Bloom Filter | Guava or Custom In-memory Filter |
| Frontend | Thymeleaf + HTML/CSS/JS |
| Build Tool | Maven or Gradle |
Prerequisites:
- Java 17+
- PostgreSQL
- Redis
- Kafka & Zookeeper
- Maven or Gradle
# Clone the project
git clone https://github.com/Shivam171/url-shortner.git
cd url-shortener
# Configure application.yml (DB, Redis, Kafka settings)
# Run the app
./mvnw spring-boot:runPRs, suggestions, and feature requests are welcome. Feel free to open issues or contribute with enhancements.
MIT License – use it freely, with credit.
I'm actively building and improving this system every week. You can follow along or contribute — let’s make a truly scalable URL shortener together.
— Built for scale. Developed with passion.