How Booking com designed and scaled their highly available and performant User Review System

Watch the video explanation ➔

The review system for Booking is one critical service as it drives business and hence making it highly available becomes extremely crucial.

Here’s a thread about building a highly available system and the architecture of Booking’s Review Service 👇‍


Review shown on are authentic and it helps people make a decision and thus bring in revenue.

The review system is also a high-throughput system as people can land on it from Search Engines or internal navigation.

Review API Service

The core Review Service will be a simple REST-based API that exposes endpoints to create, read, update, and delete reviews.

Review Service of Booking handles 10,000 requests per second at peak with p99 of 50ms.

Given that the latncy requirement is too stringent, Booking serves most of the review from a centralized remote Cache and has done a bunch of optimizations on the Database through materialized views.

Configuring storage has 250 million reviews and given the amount of info a review holds, they store these reviews in a shared datastore.

Given Booking needs storage to be highly available, they configured each database to have multiple replicas and that too across regions to tolerate regional outages.

Auto-scaling Storage

Given that the traffic for a travel business surges during the holiday season, the database needs to be scaled up to handle the load.

But keeping the database scaled up throughout the year without much load is very inefficient and hence required some so of storage autoscaling.

Scaling up and down a sharded database requires us to add and remove nodes; but doing this naively requires the data to be re-partitioned.

To keep the movement to a bare minimum during scaling, chose Consistent Hashing for determining data ownership.


  1. Review Service is a simple REST service
  2. Centralised cache and materialized views for quicker response
  3. Database is sharded to handle larger loads
  4. Data ownership is determined by Consistent Hashing
  5. Database is replicated across within and across regions for HA

Here's the video ⤵

Courses I teach

Alongside my daily work, I also teach some highly practical courses, with a no-fluff no-nonsense approach, that are designed to spark engineering curiosity and help you ace your career.

System Design Masterclass

A no-fluff masterclass that helps experienced engineers form the right intuition to design and implement highly scalable, fault-tolerant, extensible, and available systems.

Details →

System Design for Beginners

An in-depth and self-paced course for absolute beginners to become great at designing and implementing scalable, available, and extensible systems.

Details →

Redis Internals

A self-paced and hands-on course covering Redis internals - data structures, algorithms, and some core features by re-implementing them in Go.

Details →

Writings and Learnings

Knowledge Base



Arpit's Newsletter read by 90000+ engineers

Weekly essays on real-world system design, distributed systems, or a deep dive into some super-clever algorithm.