🎬 Booking Service Demo

Production-grade distributed booking system with race-condition protection using Redis distributed locks and database transactions.

🎬 Interactive Demo: This dashboard demonstrates the booking service UI with mock data. All features work - create bookings, run concurrency tests, view activity logs.

For Production: The backend uses Redis distributed locks + PostgreSQL transactions to prevent race conditions. Deployable on Cloudflare Workers (serverless) or Docker (PostgreSQL + Redis).

🏥 Service Health

🎬 Demo

Status

< 10ms

Response Time

🎬

Mode

API Endpoint: Cloudflare Pages (Demo)

➕ Create Booking

⚡ Concurrency Test

Send 10 simultaneous requests for the same time slot. Only 1 should succeed.

Success

Failed

📝 Activity Log

Title	Start Time	End Time	Status
Team Standup	-	-	CONFIRMED
Product Review	-	-	CONFIRMED

Technical Implementation

Node.js TypeScript PostgreSQL Redis Cloudflare Workers D1 Database Docker

Race-Condition Protection

Redis Distributed Locks: Uses Redlock algorithm to acquire exclusive locks on time slots before booking
Database Transactions: PostgreSQL transactions with SERIALIZABLE isolation level for conflict detection
Idempotency Keys: Prevents duplicate bookings from retry logic
Optimistic Locking: Version-based conflict detection as fallback
Audit Logging: Complete trail of all booking attempts with timestamps

Architecture Options

Cloudflare Workers (Serverless): Uses D1 database + Upstash Redis for fully serverless deployment
Docker (Traditional): PostgreSQL + Redis containerized setup for self-hosted deployment
Hybrid Mode: Demo mode with mock data for testing without backend

API Endpoints

POST /api/bookings - Create a new booking with race-condition protection
GET /api/bookings - List all bookings
GET /api/bookings/:id - Get booking by ID
DELETE /api/bookings/:id - Cancel a booking
POST /api/concurrency-test - Run concurrent booking simulation
GET /api/health - Health check endpoint

Key Features

Time slot conflict detection with sub-millisecond precision
Distributed locking prevents double-booking across multiple instances
Transaction rollback on conflict with detailed error messages
RESTful API design with proper HTTP status codes
Interactive dashboard demonstrating concurrency concepts
Comprehensive test suite with 90%+ coverage

Overview

Distributed booking system designed to handle concurrent requests safely. Implements multiple layers of race-condition protection including Redis distributed locks, database transactions, and idempotency handling. Demonstrates production-ready patterns for preventing double-booking in high-concurrency scenarios.

Key Features

Race-Condition Safe: Only 1 of 10 concurrent requests for the same time slot succeeds
Redis Distributed Locks: Redlock algorithm for cross-instance coordination
Database Transactions: PostgreSQL with SERIALIZABLE isolation for conflict detection
Idempotency: Prevents duplicate bookings from retry logic
Audit Logging: Complete trail of all booking attempts
Interactive Demo: Live concurrency test showing protection in action

Tech Stack

Node.js TypeScript PostgreSQL Redis Cloudflare Workers D1 Database Docker Upstash

← Back to Projects