Advanced strategies for building robust and scalable software architectures
Constantinov, Alexandru (2024)
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024120332356
https://urn.fi/URN:NBN:fi:amk-2024120332356
Tiivistelmä
Software systems today are becoming more complex and interconnected than before, most of them operating at an enterprise scale. This puts a higher value on design approaches that promote robustness, maintainability, and the ability to evolve over a certain period.
This thesis explores advanced strategies for designing and assessing software architectures to meet these new challenges, it focuses on two key areas: modern architectural patterns that support flexibility and scalability, and quantitative metrics for evaluating the structural quality of code. In the thesis it will be examined how concepts like microservices, event-driven architectures, and serverless computing can enable more modular, strong, and resilient systems.
To complement these patterns, a certain number of code-level metrics that provide insights into a system’s design health will be investigated. These include measures of coupling and cohesion, cyclomatic complexity, and the degree of abstraction. Also, attention will be paid to the ideas of efferent and afferent coupling and how these can relate to a module’s stability and ability to absorb changes. The end goal is to define a toolbox of strategies for creating systems that are not only functional but well-engineered.
This thesis explores advanced strategies for designing and assessing software architectures to meet these new challenges, it focuses on two key areas: modern architectural patterns that support flexibility and scalability, and quantitative metrics for evaluating the structural quality of code. In the thesis it will be examined how concepts like microservices, event-driven architectures, and serverless computing can enable more modular, strong, and resilient systems.
To complement these patterns, a certain number of code-level metrics that provide insights into a system’s design health will be investigated. These include measures of coupling and cohesion, cyclomatic complexity, and the degree of abstraction. Also, attention will be paid to the ideas of efferent and afferent coupling and how these can relate to a module’s stability and ability to absorb changes. The end goal is to define a toolbox of strategies for creating systems that are not only functional but well-engineered.