Autonomic Resource Provisioning Framework for Service-based Cloud Applications: A Queuing-Model Based Approach

Abstract

The user's request changes dynamically in service-based cloud applications, which requires optimal amount of computing resources to meet service-level agreements (SLAs). The existing server-side resource allocation mechanisms have limitations in provisioning the required resources to handle the incoming load on the basis of user's requests. To overcome the aforementioned situation, cloud computing provides ample amount of computing resources to meet the SLAs. There are possibilities that cloud resources might not be properly utilized and might suffer over and under utilization. In this study, the authors have proposed an autonomic resource allocation framework, that automatically provisions (allocate and deallocate) the required computing resources as per the load. In this study, the proposed model leverages the queuing model to optimize the resource allocation process. The primary goal of this study is to improve the virtual resource utilization and response time with respect to the existing methods. Finally, the results have shown that the response time and resource utilization have been improved.