Modernizing Enterprise Applications
A Spring Boot monolith strangled to Quarkus + Camel microservices, one seam at a time, driven by an AI Development Lifecycle (ADLC) and tested throughout.
What this book is
A single continuous shipping/e-commerce narrative: build a believable
Spring Boot monolith with six bounded contexts — order, inventory,
payment, shipping, notification, review — then strangle it to Quarkus
and Apache Camel microservices one seam at a time, in rising order of
difficulty. Each pattern (ACL, CDC, outbox, saga, CQRS, service mesh)
is introduced at the moment the migration forces it, every extraction
is proven against the monolith's own behavior-equivalence suite before
its module is decommissioned, and the whole thing is executed through
an AI Development Lifecycle (ADLC) — a disciplined
Frame → Map → Plan → Generate → Verify → Operate → Reconcile loop with
explicit human gates — that the book both teaches and demonstrates on
its own build. This project's own _plans/ ledger is
Exhibit A: the book is built the way it teaches.
Setting Up
Toolchain (SDKMAN/JDK 25/Maven/Quarkus+Camel CLI), Docker Engine and the compose observability stack, the repo, the ADLC ledger, how to run every example.
Why Modernize
Modernization as disciplined engineering; when microservices are a bad idea; strategies; the assessment rubric and evolution arc reframed for 2026.
The AI Development Lifecycle (ADLC)
What replaces the SDLC: phases, agent roles, MCP tools, human gates, the safety net, and a full worked loop on a trivial change.
The Reference Monolith
Build the Spring Boot 'before' picture: domain, modules, persistence, API, seed data, deliberate smells, and the test suite that becomes the equivalence suite.
Finding the Seams
Strategic + tactical DDD, hexagonal, event storming, coupling theory, modular monolith / decomposed DBs — deciding where to cut.
The Strangler Fig in Practice
The strangler family and the first two extractions (review, notification); ACL/content-based routing; going event-driven.
Data Across the Seam
Shared data → CDC/log-tailing (+ inventory extraction) → outbox → event-sourcing/CQRS → ACID→ACD.
Coordinating Across Services
Distributed transactions, choreographed saga (+ payment), orchestrated saga (+ shipping), failure modes and the resilience chassis.
Communication & Contracts
REST/gRPC/GraphQL (+ order extraction & gateway, strangler completes), the Quarkus/MicroProfile chassis, the schema/API/service registry.
Operating the Modernized System
Deployment patterns; service mesh, observability (OTel/LGTM/Kiali), distributed tracing — on minikube.
Delivering & Reflection
CI/CD for the strangler on GitHub Actions (the workflow files, the equivalence-gate job, the deployment jobs named explicitly), GitOps, progressive delivery, supply-chain security; then the pattern-map/chassis/anti-patterns/what's-next close.
Appendices
Reference material that supports the main narrative without interrupting it — starting with deploying the finished system to OpenShift (Helm chart, Routes, Security Context Constraints, the internal registry), verified on a real OpenShift Local cluster.