College of Graduate Studies: Theses & Dissertations

Term of Award

Spring 2026

Degree Name

Master of Science, Information Technology

Document Type and Release Option

Thesis (open access)

Copyright Statement / License for Reuse

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Department

Department of Information Technology

Committee Chair

Hayden Wimmer

Committee Member 1

Atef Mohamed Shalan

Committee Member 2

Jongyeop Kim

Abstract

This research aims to investigate performance optimization and reliability issues related to cloud-based computing environments through an analysis of three key infrastructure components: virtualized CPU resource management, distributed API rate limiting, and web server deployment architectures. This research combines machine learning and system experimentation as a way of exploring the impact of infrastructure-level behaviors on overall system performance and scalability. The first component of the research focuses on analyzing CPU Fragmentation in Virtualized Environments, where unbalanced workload allocation on Virtual CPU Cores causes increased tail latency, resulting in Service Level Agreement violations. Metrics are analyzed using the Random Forest classifier on a 6vCPU Virtual Machine running latency-sensitive microservice applications. The second component of the research focuses on analyzing the performance of distributed API Rate Limiting functionalities, where the mechanism is implemented at the Load Balancer Level, using Redis as the shared state storage. Experiments are conducted on various types of traffic, including stationary, bursty, flash crowd, and non-stationary traffic, to analyze the performance of the Rate Limiting mechanism, including the accuracy, responsiveness, and sensitivity of the parameters used. The third component of the research focuses on analyzing the performance of Apache and Nginx Web Servers, running in Docker Containers and Virtual Machines, using Round Robin and Least Connections Load Balancing algorithms. Traffic Workloads are simulated using the Siege tool. Experimental results demonstrate that infrastructure-level configuration choices and workload dynamics significantly influence system stability, enforcement accuracy, and performance efficiency. Machine learning based resource management improves workload balance and latency predictability in virtualized systems, while realistic traffic modeling reveals important behavioral deviations in distributed rate-limiting mechanisms. Additionally, web server performance varies depending on deployment environment and load-balancing strategy, highlighting trade-offs between scalability, isolation, and responsiveness. The findings will also offer practical insights for system architects and infrastructure engineers aiming to optimize cloud-native systems, virtualized infrastructure, and web services platforms. By integrating machine learning-based resource management with the empirical evaluation of infrastructure, this research will contribute to the betterment of the predictability and resiliency of the performance of distributed computing systems.

INDEX WORDS: Cloud computing, Machine learning, Random Forest, CPU fragmentation, Virtual machines, Distributed rate limiting, Redis, API gateways, Traffic patterns, Web servers, Apache, Nginx, Docker containers, Load balancing, Performance evaluation, Cloud infrastructure.

Research Data and Supplementary Material

No

Download

Share

COinS