Skip to content

Level 4 · Organizing, Migrating & Operating Microservices

Level 4 zooms out from any single service to the system and the organization around it: how you migrate a real monolith without a big-bang rewrite, how teams and service boundaries pull on each other, and — just as important — when the honest answer is not to split at all.

What was run, and what wasn't

This level runs a real strangler-fig migration against a small real monolith (a Flask app), routing selected paths to a new extracted service while the rest keeps serving from the monolith, plus real blue-green and canary deployment scripts against local services. Organization-level claims (team topologies, Conway's Law) are presented as documented industry patterns, not something a single-laptop course can run — that distinction is kept explicit throughout.

Prerequisites: Levels 1–3.

Modules

  1. The Strangler Fig Pattern: Migrating a Monolith Incrementally: a real reverse proxy peeling one route at a time off a real monolith.
  2. Team Topologies & Conway's Law: why your service boundaries tend to mirror your org chart, whether you plan it or not.
  3. Versioning & Backward Compatibility Across Services: shipping a breaking API change without a synchronized deploy.
  4. Shared Libraries vs Code Duplication: the real coupling cost of a shared internal library, and when duplication is cheaper.
  5. Deployment Strategies for Many Independent Services: real rolling, blue-green, and canary deploy scripts against local services.
  6. Observability Across Service Boundaries: correlating real logs, metrics, and traces from three services into one incident timeline.
  7. Testing Strategies in a Microservices World: where unit, contract, integration, and end-to-end tests each actually earn their cost.
  8. Anti-Patterns: the Distributed Monolith & Chatty Services: a real example of "microservices" that can't be deployed independently, and why.
  9. When Not to Use Microservices: the honest cases where a monolith is the better engineering decision.
  10. Capstone: Build, Break & Observe a Real Multi-Service System: a multi-service system built from scratch, with a real injected failure and a real trace showing exactly where it broke.

After Level 4

You've decomposed a real system, kept its data consistent under real failure, made it resilient, and migrated a real monolith into it — the same loop you'd run on a production system, just at a scale you can finish on a laptop.