The Power of Lacinia & Hystrix in Production
summary
Combines Lacinia resolvers with Hystrix def-command wrappers to make GraphQL APIs fail fast and recover gracefully; example: adding image-parameter resolvers for adaptive images and wrapping each external call in a Hystrix command. Small 100ms latency variance yields ~6% users hitting 500ms, but one dependency spiking to 2s produces 100% seeing ~20s and 70% lost requests; Hystrix flattens the tail during load, provides fallback strategies and cached fallbacks, and exposes function-level metrics via its dashboard so the approach both improves latency tails and survives partial failures.
AI-written summary · about
chapters (11)
- 0:00Intro
- 5:08Lacinia Example
- 8:47GraphQL
- 12:32Moving Forward
- 13:52Extending the Model
- 17:28Dealing with Bad Times
- 19:29The Source of Evil
- 20:35The Happy Path
- 26:43In Practice
- 31:43Conclusion
- 36:37Questions
related talks
43mFault tolerance – look, it’s possible!
39mGraphQL Accelerated
54mBattle of the Circuit Breakers: Hystrix vs. Istio
24mA GraphQL Survival Kit
37mCaching Deep Dive: The Ultimate Way To Speed up Your GraphQL API
25mOptimizing GraphQL Applications
25mGrafast: A Declarative Solution To GraphQL's Execution Woes
23mHow to edge cache GraphQL APIs
