Reading an RFC Trade-offs Section — Comprehension Exercises

RFC-0047: Trade-offs Analysis (Section 5) Two integration approaches were evaluated for inter-service communication after extraction. Approach A — Event-driven (async messaging via a message broker): Services publish events; downstream services subscribe and react. This approach provides high availability: the publisher does not wait for consumers, so a slow or unavailable consumer does not block the publisher. However, it introduces eventual consistency — consumers may lag behind the publisher, meaning the system is temporarily in an inconsistent state. Debugging distributed event flows requires additional tooling (distributed tracing, dead-letter queues). Operational complexity is high. Approach B — Request-response (synchronous HTTP/gRPC): Services call each other directly and wait for a response. This provides strong consistency — the caller immediately knows whether the operation succeeded. However, it creates tight coupling: if the billing service is slow, the checkout service blocks and may time out. Cascading failures are a real risk at scale. Operational complexity is lower. Recommendation: We recommend Approach B for the initial extraction. Our team has no existing message broker infrastructure and limited experience operating one. The three services being extracted have relatively low interdependency, making synchronous calls manageable. We will revisit this decision at scale.