Design a real-time chat system with hot groups
Quick Overview
This question evaluates system design and real-time distributed systems skills, including designing scalable chat architectures that support 1:1 and group conversations, WebSocket-based delivery and routing across multiple gateway servers, message persistence and history sync, multi-device consistency and read receipts, plus the ability to define, detect and optimize for the Top N "hot" group chats. Commonly asked to assess reasoning about high-level architecture, concrete data schema and indexing, routing and state synchronization, and availability/disaster-recovery trade-offs, it tests both conceptual understanding and practical application by requiring a clear architecture together with detailed data model and scalability/performance considerations.
Design a real-time chat system (similar to an in-app messaging feature) that supports:
- 1:1 and group conversations
- Real-time delivery over WebSockets (or equivalent)
- Message persistence and history sync
- Read receipts (at least per-user “last read”)
- Multi-device users (same user logged in on multiple clients)
- High availability / disaster recovery considerations
Additional requirement:
- The system must optimize for the Top N “hottest” group chats (e.g., groups with extremely high message throughput and/or many concurrently online participants). Explain what “hot” means and how you detect it.
The interviewer expects particular attention to:
- A clear high-level architecture
- A concrete data schema (tables/collections, keys, indexes)
- How messages get routed when you have multiple WebSocket gateway servers
- Scalability and performance trade-offs