Microservices — Roadmap
When to split a monolith, gRPC between services, service discovery with Consul.
Real-World Analogy
A small restaurant vs a food court: one restaurant (monolith) has one kitchen, one bill, one staff — easy to run until the menu explodes and the kitchen can’t keep up. A food court (microservices) has specialized stalls that can open and close independently, scale on their own, and fail without closing the whole court. The trade-off is coordination: who takes the order, and how does sushi know the stir-fry is ready?
What you will learn
Microservices are not a default — they’re a trade-off. This track covers the real costs of splitting (distributed transactions, network calls, operational overhead), how to find the right seams before cutting, and the infrastructure that holds services together: gRPC contracts, service discovery, the API gateway pattern, and the reliability patterns (circuit breakers, bulkheads, retries) that prevent one slow service from cascading into a full outage.
Chapters in this track
- Monolith vs Microservices — when to split, the strangler fig pattern, modular monolith as the middle ground
- gRPC Between Services — Protocol Buffers, code generation, streaming, error handling, schema evolution
- Service Discovery — DNS-based discovery, Consul, client-side vs server-side LB, service mesh
- API Gateway — routing, auth, rate limiting, request transformation, what not to put in the gateway
- Inter-Service Reliability — timeouts, retries, circuit breakers, bulkheads, hedged requests