GraphQL Subscriptions - Scaling real-time in the real world
summary
GraphQL Subscriptions are in the official spec and are usable in production today, with working examples like WhatsApp-Server-GraphQL-Live-RxJS-Meteor and an Angular frontend plus graphql-code-generator; implementation uses WebSockets with fallbacks, RxJS observables for live queries, and pub/sub backbones (Redis or Kafka) to scale and handle QoS. Practical notes: separate stateful connection handling from stateless GraphQL servers, failover and offline sync via Apollo/Relay patterns, and new hooks like MongoDB 3.6 change streams for reactive backends.
AI-written summary · about
chapters (25)
- 0:00Introduction
- 0:48What is GraphQL Subscriptions
- 2:26How does GraphQL work
- 6:02Tomorrows Talk App
- 6:17GraphQL Code
- 7:02GraphQL RealTime
- 7:37Types of RealTime
- 8:40Magic in the Cloud
- 10:03What if I dont have Reactive
- 10:35Polling
- 11:41Realtime
- 12:07PubSub
- 13:53The Client
- 15:20The GraphQL Subscription
- 19:03Common Problems
- 20:49Bidirectional Transport
- 21:23WebSocket
- 22:52Offline
- 23:16Local Store
- 23:52Scaling
- 25:55Pubsubs
- 27:33Live queries
- 29:50GraphQL Rxjs
- 33:00Libraries
- 33:16Request from Service Providers
related talks
34mGraphQL real-time subscriptions
15mGraphQL Subscriptions in Production Is Easy Isnt It?
30mBest Practices for GraphQL and GraphQL Subscriptions at Scale
52mBuilding a Realtime Chat with GraphQL Subscriptions
24mRealtime React Apps with GraphQL
18mCrafting compelling real-time web experiences with GraphQL
32mGraphQL Future
22mReal-Time Data Updates for Neo4j Using GraphQL Subscriptions