How to Design Large-Scale Systems: From Business to Technical Architecture

Background

Architecture design includes two categories: large systems and simple systems. While many articles discuss simple system design, large‑scale system design is less covered and is the focus of this article.

Differences Between Large and Simple Systems

System type: Large systems are distributed; simple systems are usually monolithic.

Business complexity: Large – complex; Simple – simple.

Scale complexity: Large – complex; Simple – simple.

Technical complexity: Large – complex; Simple – simple.

Resource investment: Large – high; Simple – low.

Cross‑departmental: Large – yes; Simple – no.

Is Large‑System Design Just Distributed Design?

In plain terms, a large system consists of multiple monolithic or simple systems and is typically distributed. Design should start from business concerns before jumping to technical solutions; otherwise, architecture becomes a mere stack of technologies.

How to Design a Large System

Apply the “divide‑and‑conquer” principle: break the big problem into smaller, manageable parts. System design follows three layers: business, system, and technology.

Business layer: Clarify the problem to be solved.

System layer: Create the overall system blueprint.

Technology layer: Choose implementation technologies.

Design Steps

The core steps mirror enterprise architecture: business architecture, system architecture, and technology architecture, which together form a complete solution.

Business Architecture

Business architecture defines the governance structure, capabilities, processes, and data that support the business.

Governance structure (organization, channels, partners)

Business capabilities (value chain, functional domains)

Business processes (core, branch, rules)

Business data (domains, models, rules)

Application Architecture

Application architecture describes a set of application systems and their interactions, each representing a logical functional group that supports business functions and manages data assets.

Identify required business and data functions.

Map functions to application components.

Design interaction and collaboration between applications.

Data Architecture

Data architecture provides a blueprint for managing data assets, covering data classification, logical and physical models, and data management.

Technical Architecture

Technical architecture defines the technology stack, including technology requirements, selection, architectural design, and management.

Technology selection: platforms, frameworks, middleware, hardware, network.

Architecture design: component, network, deployment.

Technology management: standards, case studies.

Solution

The complete solution integrates business, application, data, and technical architectures into two documents: a Business Architecture document and a Technical Architecture document (which includes application, data, and technical sections).