Using Apollo Federation in a microservice architecture in Kubernetes
summary
Demo stitches separate GraphQL services using Apollo Federation with extend/@key/external fields to resolve cross-service fields, deployed as independent Node servers on minikube. A gateway composes accounts reviews products inventory, and a horizontal pod autoscaler (HPA v2 beta2) spins replicas up during load. An Artillery load test runs 120 requests with 19 timeouts locally, CPU spikes ~150 millicores, pods scale to four replicas and timeouts clear; local limit ≈2 req/s but cloud scale is unconstrained. Dynatrace monitoring and HPA behavior/policy settings are used to control scale up/down.
AI-written summary · about
chapters (6)
- 0:00Start
- 4:33Resolvers
- 5:30Schema Stitching
- 14:29Kubernetes
- 17:17Horizontal Autoscaler
- 27:33Takeaways from the Session
related talks
28mBuilding a distributed graph w/ Apollo Federation
48mApollo Federation: Connecting GraphQL Microservices
36mGraphQL Federation - The @key to Your Platform API Strategy
37mGraphQL and Microservices or How we learned to be big and small at the same time
25mStop mocking - develop frontends with real K8s setups
19mKeynote: Kubernetes Federation
06mEscape the GraphQL Monolith
30mDeep Learnings Migrating legacy services to Federated GraphQL