Product Catalog Service – High-Performance Listings Backend

The Product Catalog Service powers the creation, storage, search, and retrieval of millions of marketplace listings.
It is designed as a low-latency, horizontally scalable microservice optimized for high-read workloads typical in e-commerce and classifieds platforms.

I worked on designing, building, and optimizing this system for production-scale usage, ensuring fast listing fetches, real-time updates, and efficient search performance.

Description

The catalog service manages the complete lifecycle of product listings, including:

  • Creating & updating listings
  • Category & attribute management
  • Search indexing & synchronization
  • Enrichment (pricing, images, metadata, badges)
  • Real-time availability status
  • SEO-friendly content generation
  • Integrations with recommendations, ads, and user services

The service handles millions of read requests per minute, especially during peak traffic.

Key Capabilities Delivered

1. Listing Creation & Update API

  • CRUD APIs with strong validation
  • Optimized database writes with batching
  • Category-specific schema & dynamic attributes
  • Idempotent operations

2. High-Performance Read APIs

  • Ultra-fast listing fetch (<20ms)
  • Redis caching layer reducing DB load by 60–70%
  • Category-level & homepage caching
  • Cache invalidation on listing update

3. Search & Indexing Pipeline

  • Elasticsearch integration for full-text search
  • Real-time sync using Kafka event streams
  • Re-indexing workflows for schema changes
  • Faceted search (filters, sort, pagination)
  • Geo-based queries for local discovery

4. Media & Metadata Integration

  • Seamless integration with image/video services
  • Auto-thumbnail mapping
  • Badging engine (Featured, Verified, Urgent)
  • SEO keyword generation

5. Performance & Resilience Enhancements

  • Read/write separation with MySQL replicas
  • Circuit breakers & rate limiting
  • Async enrichment pipeline
  • Blue/green deployments
  • Observability: metrics, logs, traces

Impact

  • Reduced API latency from 90ms → 18ms through caching & query optimization
  • Increased catalog throughput by 3× via read-write splitting
  • Faster search indexing with <1 sec sync using Kafka
  • Improved homepage load speed by 30%
  • Delivered a highly stable and scalable system able to handle millions of daily active users

Technology Stack

  • Backend: Java, Kotlin, Spring Boot
  • Database: MySQL (primary + replicas), Elasticsearch
  • Caching: Redis (hot data, listings, homepage)
  • Messaging: Kafka for real-time sync
  • Infra: Docker, Kubernetes, Helm
  • Search: Elasticsearch, Kibana
  • Monitoring: Prometheus, Grafana, ELK
  • Other:
    • Resilience4j
    • Centralized config service
    • Nginx API gateway
    • OpenAPI/Swagger

Setting Up (Generic Template)

```bash git clone https://github.com/rohitshakya cd product-catalog-service/ ./gradlew build