Inside the Decision: Why Groww Chose NATS Over WebSockets

on YouTube ↗25m

" summary

Groww replaces Gorilla WebSocket + Kafka with NATS for live prices, cutting backend complexity and giving subjects for per-user feeds so millions of traders get individualized streams without creating Kafka topics. NATS uses a fire‑and‑forget model and built‑in reconnection for near‑zero latency and no retention delays; resource usage handles 2.5x more users for the same machines. Also replaces polling for order status, and uses NATS request‑reply for HNS messaging.

AI-written summary · about

" chapters (7)

1. 0:00Introduction
2. 3:15Problems with Kafka and WebSockets
3. 9:58Why NATS was chosen
4. 12:51NATS features for live prices
5. 20:09NATS benefits and results
6. 22:56Other NATS use cases
7. 24:46Conclusion and Q&A

" related talks

• 37mIntro: NATSWaldemar Quevedo · KubeCon + CloudNativeCon Europe 2019
• 37mConnect All the Things! Using NATS To Simplify Communication EverywhereJeremy Saenz, Todd Beets · KubeCon + CloudNativeCon North America 2022
• 40mThe Zen of High Performance Messaging with NATSWaldemar Quevedo Salinas · Strange Loop 2016
• 29mA New Generation of NATSMatthias Hanel, Jean-Noël Moyne · KubeCon + CloudNativeCon North America 2021
• 21mNATS: A high-performance messaging system for communication between servicesDoni Rubiagatra · GopherCon Vietnam 2019
• 25mBeyond Kafka: Leveraging Event-Driven Architectures with Go and NATS JetStreamLudovico Russo · GoLab 2024
• 35mThe NATS Stack - Libraries Extensions and the Execution EngineTomasz Pietrek, Jordan Rash · KubeCon + CloudNativeCon Europe 2025
• 34mState of NATS: Core, Persistence, & Materialized ViewsTomasz Pietrek · KubeCon + CloudNativeCon Europe 2022