/url-shortener

Scalable URL Shortener Backend with High Availability and Clean Architecture

Primary LanguageGo

Scalable URL Shortener Backend with High Availability

This project is the backend component of a scalable URL shortener with high availability, implemented using Docker, Go, MongoDB, ZooKeeper, and Redis. It provides a reliable and efficient backend system for shortening long URLs into shorter, more manageable ones, while ensuring high availability and fault tolerance.

Background

Learn to build systems with production standards that have high scalability. Based on the case study learned about the url shortener in the following article: System Design URL Shortening Service. I tried to build a shorterner url system following the troubleshooting based on the guidelines in the article.

Features

  • Shorten long URLs into compact, easy-to-share links.
  • High availability design to ensure system reliability.
  • Scalable architecture with distributed databases and caching.
  • Fault tolerance mechanisms for continuous operation.

Prerequisites

Make sure you have the following prerequisites installed:

  • Docker (version >= 23.0.1)
  • Docker Compose (version >= v2.16.0)

Configuration

Create a .env or replace .env.example / .env.example.replicated file in the project directory:

PORT=4001

DATABASE_USER=aryahmph
DATABASE_PASSWORD=aryahmph
DATABASE_HOSTS_PORTS=mongo1:27017
DATABASE_REPLICA_SET=-
DATABASE_NAME=url_shortener
DATABASE_MIN_POOL_SIZE=10
DATABASE_MAX_POOL_SIZE=100

ZOOKEEPER_HOSTS=zoo

REDIS_ADDRESS=redis:6379

Installation

  1. Clone the repository:
git clone https://github.com/aryahmph/url-shortener.git
  1. Change to the project directory:
cd url-shortener
  1. Start the Docker containers using Docker Compose:
make standalone-up
  1. If you run replicated mode, run this command:
make replicated-up
docker compose exec mongo1 /scripts/setup.sh
  1. Connect to http://localhost:4001/api/v1

Usage

The backend component provides the API endpoints for URL shortening and redirection. To interact with the backend, you will need to set up a frontend or use an API testing tool such as Postman.

Import postman collection file to your Postman.

Testing Result

This section provides an overview of the testing results obtained using [K6] (https://github.com/grafana/k6), a popular open-source load testing tool.

Run script.js file with K6

k6 run testing/script.js

Test Scenario

  1. Scenario 1 : Link Creation and Get Link
  • Simulates create and get link process with 1 : 10 ratio.
  • Verifies the correctness of create and get link functionality.
  • Measures response times for create and get link requests.

Test Environment

The load tests were conducted in the following environment:

  • Operating System: Ubuntu 20.04.5 LTS
  • CPU: Intel® Core™ i5-8265U CPU @ 1.60GHz × 8
  • RAM: 12GB

Results Summary

Scenario 1: Link Creation and Get Link (Standalone each service)

  • Virtual Users: 1000
  • Duration : 30s
  • Total Requests: 330011
  • Requests per Second (RPS): 10979/s
  • Average Response Time: 813.76µs
  • 90th Percentile: 1.17ms
  • Error Rate: 0%

Todo List

  • Add support for custom URL aliases and expiration of shortened URLs.
  • Migrate from docker compose to Kubernetes (Still learning).
  • Implement load balancing and horizontal scaling for distributing traffic across multiple backend instances.
  • Implement Redis clustering for high availability and fault tolerance of caching.
  • Implement Unit & Integration Testing
  • Improve caching strategy.
  • Improve performance testing strategy.

Contributing

Contributions are welcome! If you would like to contribute to this project, please follow these steps:

  1. Fork the repository.
  2. Create a new branch.
  3. Make your changes.
  4. Commit your changes.
  5. Push your branch to your forked repository.
  6. Open a pull request.