Mastering System Design: From Core Concepts to Practical Architecture

System Definition

A system is a collection of entities and the relationships between them whose combined functionality exceeds the sum of its parts. The article focuses on artificial systems that require human involvement.

System = a set of entities and the relationships between them, with functionality greater than the sum of individual entities.

System Analysis

System analysis studies the structure, interfaces, behavior, and constraints of a system to improve decision‑making and overall performance. Recommended analysis approaches include taxonomy, hierarchical decomposition, logical relationships, top‑down, bottom‑up, and external‑to‑internal perspectives.

System analysis aims to research the interactions of system components, external interfaces, overall behavior, and limitations to provide references for future changes and decisions.

Modeling Foundations

Modeling transforms vague requirements into concrete representations. Key model types:

Scientific model – mathematical or simulation based.

Concept model – abstract ideas describing system operation.

Business model – captures processes and organization.

Domain model – essential entities, rules, and structures.

System model – overall structure and architecture.

Analysis, design, and physical models – refine and implement solutions.

Use‑Case Driven Modeling

Start from stakeholder expectations, define actors , scenarios , and use cases (actor + action + object). Example: an analyst creates reports for customers. The process expands to cover roles, permissions, and business processes, then derives business, concept, and system models.

Domain‑Driven Modeling

Begin with a top‑down view, involve domain experts to extract essential entities, aggregates, factories, and repositories. A typical layered architecture consists of:

Interface layer – presents information and interprets user commands.

Application layer – thin coordination layer, no business logic.

Domain layer – core business logic, entities, value objects, domain events, repositories.

Infrastructure layer – technical support, persistence, external integrations.

Architecture Derivation

Architecture is an abstract description of entity relationships that guides design. Classification includes layered, event‑driven, micro‑kernel, micro‑service, and cloud‑native architectures. Derivation combines:

Deduction – abstract use cases into business models, then add technical concerns.

Induction – group and generalize requirements.

Typical architecture output items:

Solution overview

Design constraints

Technical selection

System structure (deployment diagram)

Key technical designs (security, algorithms)

Interface design (protocols, data structures)

Data design (SQL/NoSQL, ER models)

Quality prediction

Design Principles and Patterns

Core principles:

GRASP (responsibility assignment)

SOLID – OCP, SRP, LSP, ISP, DIP

Law of Demeter (least knowledge)

High cohesion, low coupling

Common design patterns:

Creational – Factory, Builder, Singleton, Prototype, Abstract Factory

Structural – Facade, Adapter, Proxy, Decorator, Bridge, Composite, Flyweight

Behavioral – Observer, Strategy, Command, State, Template Method, Chain of Responsibility, Mediator, Visitor

Design Tools and Modeling Languages

Effective modeling relies on standardized notations. The most widely used is UML (Unified Modeling Language), complemented by SysML or OPM when needed. Core UML elements include participants, use cases, boundaries, business entities, packages, classes, relationships, components, and nodes. Views are divided into:

Static – use‑case diagram, class diagram, package diagram.

Dynamic – activity diagram, state diagram, sequence diagram, collaboration diagram, swim‑lane diagram.

Popular tooling:

draw.io

StarUML

Visual Paradigm

Requirement Analysis

Architects must clarify:

Problem domain and legality.

Target audience and beneficiaries.

Business value and goals.

Prioritized, verifiable, and uniquely identified requirements.

Scope, boundaries, and external interactions.

Key questions include who the customers are, what value the system delivers, constraints on cost, time, safety, and quality.

Model Establishment Process

1. Entity analysis – ask “What is the system?”, “What are its components?”, “How are they related?”, “What is the boundary?” and “What are the usage scenarios?”.

2. Relationship analysis – identify topology, connectivity, address, sequence, membership, ownership, and interpersonal relations.

3. Functional analysis – represent functions as subject + action + object (e.g., “analyst creates report”).

Model Types and Methods

Model classifications:

Business model

Concept model

System model

Analysis model

Design model

Physical model

Common modeling methods:

Domain‑Driven Design (DDD)

Use‑Case Driven (UDD)

Four‑color modeling

CRC (Class‑Responsibility‑Collaborator)

CQRS (Command Query Responsibility Segregation)

Architecture Output and Summary

Architecture documents should contain:

Problem definition and prioritization (primary, secondary, urgent, non‑urgent).

Deductive and inductive reasoning results.

Bottom‑up aggregation of use cases into business models, then into system models.

Top‑down abstraction of client needs into high‑level solutions.

The final architecture balances both bottom‑up and top‑down approaches, ensuring that the system meets stakeholder benefits while remaining technically feasible.

Design Constraints and Guidelines

Model constraints must respect industry‑proven principles such as:

GRASP responsibilities

Information expert, creator, low coupling, high cohesion

Controller, polymorphism, pure fabrication, indirection, protected variations

Design principles (SOLID, DIP, ISP, LSP, OCP) and patterns should be applied to achieve extensibility, maintainability, and testability.

Conclusion

System design is analogous to medical diagnosis: understand the problem, define the “medicine” (principles, patterns, tools), and prescribe a solution. Mastery requires a solid mental model, systematic analysis, and disciplined application of modeling, architecture derivation, and design principles.