Architecture Methodology: From Problem Identification to Application Logic Design

The article begins with a background on recurring architectural failures caused by inadequate iteration methods and emphasizes the need for a systematic approach to architecture.

Problem Identification : It highlights that each new requirement often leads to a complete redesign, illustrated by the code snippet 需求分析，架构实现，（新需求，架构改动）* n = 推倒重来。 .

What Is Architecture? The ISO/IEC 42010 definition is quoted: The fundamental organization of a system, embodied in its components, their relationships to each other and the environment, and the principles governing its design and evolution. This definition underpins the subsequent discussion.

Architecture Classification : Six categories are introduced—product functional architecture, business capability architecture, application logical architecture, application physical (deployment) architecture, infrastructure architecture, and others—each with distinct purpose, audience, and content.

Ability vs. Responsibility : The article differentiates product capabilities (what a product can do) from architectural responsibilities (how internal modules fulfill those capabilities).

Derivation Approaches : Two main strategies are described—top‑down (starting from a known problem definition) and bottom‑up (deriving from concrete use‑cases). The bottom‑up path is emphasized for its practicality when domain knowledge is limited.

Model Layers : Three modeling layers are defined—business concept model (problem space), system model (solution space), and storage model (physical data). These layers guide the transition from analysis to design.

Bottom‑Up Derivation Steps : The process includes extracting a use‑case set, deriving a concept model, performing inductive grouping based on cohesion and coupling, and refining modules using naming, cohesion metrics, creator and information‑expert principles.

Business Process Integration : Business flows are incorporated to identify module boundaries and handle exception branches, ensuring comprehensive coverage.

Logical Architecture Construction : The logical architecture is built by merging business concept architecture, system models, and processes, then refining with performance, stability, and cost considerations (e.g., intelligent cube for low‑latency reporting).

Non‑Functional Modules : Techniques for deriving modules driven by performance, stability, and cost are discussed, including leveraging open‑source solutions, academic research, and industry expertise.

Summary of Derivation Paths : Four sub‑paths—business concept architecture, system model, system process, and non‑functional support—combine to form a complete logical architecture.