Overcoming Service Communication Issues and Challenges
summary
Compares strict vs loose contracts and recommends always versioning strict schemas from the start, using vendor mime types for contract changes and endpoint versioning only for new functionality; shows stamp coupling and bandwidth pain with a concrete example (500 KB profile payload → 1 GB/s vs 400 KB/s when called 2,000×/s), explores field selectors, new endpoints, GraphQL, messaging (private east‑west APIs with fitness functions), gRPC with protocol buffers (≈10× faster, ~7–8 ms latency) and Pact.io for contract testing.
AI-written summary · about
chapters (9)
- 0:00Introduction to service communication
- 1:36Contract coupling dimensions
- 3:01Strict versus loose contracts
- 17:30Contract versioning strategies
- 33:15Stamp coupling and bandwidth issues
- 40:02Protocol considerations for communication
- 47:22gRPC for low-latency calls
- 50:10Guidelines for choosing protocols
- 53:01Q&A and closing remarks
related talks
20mService-to-service GraphQL: The New Sweet Spot!
47mBeyond REST - Contract Testing in the Age of gRPC, Kafka & GraphQL
43mEnhancing Backend Microservices Ecosystem with Contract Testing
59mUsing Pact to Deploy Microservice with Confidence
38mMicroservices, a bittersweet symphony
24mHow to test microservices and stay sane
22mMicro Testing Pains
01h 21mMicro-Service Delivery Without the Pitfalls