← Back to Documents

Project Title	FractalX — Static Decomposition Framework for Spring Boot
Document Type	Project Proposal
Version	1.0 (Sample)
Institution	SLIIT — Sri Lanka Institute of Information Technology
Department	Department of Computing
Submitted to	Dr. Supervisor Name
Date	2023 Q3
Team Members	Member 1, Member 2, Member 3, Member 4

1. Introduction

Microservices architecture has emerged as the dominant pattern for building scalable, resilient, and independently deployable software systems. However, migrating existing monolithic applications to microservices remains a significant engineering challenge, requiring weeks of manual effort, boilerplate code writing, and architectural decision-making.

This proposal presents FractalX — an open-source static decomposition framework for Spring Boot that automates the conversion of modular monolithic applications into production-ready microservice platforms through AST analysis and build-time code generation.

2. Research Problem

Existing tools for microservices decomposition (IBM Mono2Micro, ServiceCutter) rely on runtime call graph analysis or manual domain modelling, and produce recommendations rather than executable code. No existing tool provides:

• Annotation-driven, fully automated, static decomposition of Spring Boot applications
• Complete code generation including business services, infrastructure, and Docker deployment
• Automatic inter-service communication scaffolding (gRPC stubs)
• Distributed saga orchestration from source-level annotations

3. Proposed Solution — FractalX

FractalX introduces three annotations that developers add to their monolith, and a single Maven command that generates the complete microservice platform:

• @DecomposableModule — marks a module boundary
• @DistributedSaga — marks a distributed transaction method
• @ServiceBoundary — marks a cross-module interface

4. Research Objectives

1. Design and implement a static, annotation-driven AST analysis engine
2. Build a code generation pipeline producing complete Spring Boot microservices
3. Implement NetScope — an auto-generated gRPC inter-service communication fabric
4. Automate distributed saga orchestration from @DistributedSaga annotations
5. Inject observability, resilience, and Docker infrastructure automatically
6. Validate against real-world monolith scenarios

5. Methodology

The project follows an iterative build-measure-validate methodology across six phases: domain analysis, AST engine development, code generation pipeline, NetScope and saga implementation, validation, and release.

6. Expected Outcomes

• A working FractalX Maven plugin published to Maven Central
• Complete documentation (developer guide and evaluation report)
• Validated decomposition of at least 3 real-world monolith scenarios
• Open-source release under Apache 2.0 license

7. Technology Stack

Java 21, Spring Boot 3.2+/4.x, Apache Maven, JavaParser (AST), gRPC/Protobuf, OpenTelemetry, Resilience4j, Flyway, Docker/Compose, Maven Central.