Lesson 3: Architecture Styles and Quality Attributes

Architecture style is the general way a system is organized. Different styles solve different problems, and no single style is always best.

A layered architecture separates presentation, business logic, and data access. It is common in academic examples because it is easy to understand and maintain. A client-server architecture separates service providers from service consumers. A service-oriented or microservice-based architecture splits functionality into services that communicate through interfaces. An event-driven architecture reacts to events such as "payment received" or "order shipped."

Here is a practical comparison:

In a BSIT exam, you are often not asked to define styles only. You are asked to select an appropriate style for a scenario. That selection should always depend on requirements.

A system should not only "work." It should also work in a way that fits organizational expectations. These expectations are called quality attributes or non-functional requirements.

Common attributes in systems integration include:

• Performance – How fast the system responds
• Scalability – How well it handles growth
• Reliability – How consistently it works
• Availability – How often it is operational
• Security – How well it protects data and services
• Maintainability – How easy it is to update and fix
• Interoperability – How well it works with other systems

These attributes often compete with one another. A very secure system may add more steps and slow access. A highly distributed microservice design may improve scalability but make monitoring harder. A simple direct connection may be fast to build but difficult to maintain later.

This is why architecture involves trade-offs. Good designers do not claim to maximize everything. They decide which qualities matter most for the given case.

For example, an online payment integration may prioritize security and reliability first. A data dashboard may prioritize timeliness and usability. A university enrollment system during peak registration may prioritize availability and scalability.