Cloud computing: Assessing the impact of application architecture on the performance of cloud-based applications

Abstract

Cloud Computing, which involves over-the-Internet provision of dynamically scalable and often virtualized computing resources has very quickly become one of the hottest topics for practicing engineers and academics in domains related to engineering, science and art for building large-scale networks and Internet applications. The goal of this research was to investigate how application architecture impacts the performance of applications in a cloud computing environment. The specific objectives of the study were one: to identify factors driving the adoption of cloud computing for delivery computing services; two: to discover architectures used for the development of cloud based applications; three: to determine the correlation between throughput and scalability of applications and finally, to determine the moderating effect of architecture to the relationship between load and the performance of applications in a cloud computing environment. Through a detailed literature search and review, both historical and current perspective of cloud computing were examined. A conceptual framework for the research was development based on the Gartner Conceptual Framework for Application Performance Management. The experimental research methodology was adopted for the study. Microsoft Azure cloud platform and Microsoft Visual Studio Team Services was used to conduct graduated load performance tests for a convenience sample of web based applications. Data analysis was conducted by using the Pearson product-moment correlation coefficient and moderation multiple regression analysis. The literature search and review findings concurred with the observations made by that there has been limited academic research in this area of study. The findings of the study showed that there was a strong positive correlation between throughput and scalability of applications which was statistically significant, therefore the alternative hypothesis was accepted. On the other hand, the results showed that while there was a positive moderating effect of architecture on the relationship between load and performance, the moderating effect was not statistically significant, hence the null hypothesis was accepted.