Repository for Bachelor Thesis by Clemens Marx
This repository contains all code, setup and architecture diagrams related to my bachelor thesis, which evaluates monolithic and microservices architecture styles through a custom-developed public transportation application. The application implements core functionalities including route calculation, disruption reporting, and arrival time information.
Both architectures were containerized and deployed on Google Cloud using Google Kubernetes Engine. The microservices version decomposes the monolith into independent services, with special focus on the routing component as a compute-heavy bottleneck.
Performance and scalability were evaluated using load tests with Locust, collecting metrics such as response time, throughput, CPU/memory usage, bandwidth, and cloud costs. The results show that monolithic architectures perform well under low load, while microservices scale better under higher demand, at the cost of slightly higher complexity and operational overhead.
diagrams_architecture/- Architecture diagrams in PDFlocust/- Locust load testing filesmicroservice_architecture/- Microservices implementation with multiple Django servicesmicroservice_kubernetes/- Microservices deployment on Kubernetes (local and GKE)monolith_kubernetes/- Monolithic deployment on Kubernetes (as single Pod for good comparability)monolithic_architecture/- Monolithic application implementationmonolithic_cloudrun/- Monolithic deployment on Google Cloud Run (for experimentation; not included in the final thesis)