Behavior Analysis of Multifarious Cloud Applications

Cloud computing is driving a technology revolution forcing enterprises and research community to transform their existing processes into the cloud. It provides immense flexibility in IT infrastructure and aims to speed up the development and execution of business operations with immediate scaling and lucrative cost savings. Multifarious cloud applications promulgate stringent Quality of Service (QoS) requirements with disparate behavior and resource utilizations in response to fluctuating workloads. This paper presents a uniform terminology representing cloud applications, resource infrastructure and QoS criteria. Further, benchmark applications are used to analyze the behavior of applications on cloud testbed. Based on the measurements obtained, we derive patterns of resource usage using statistical techniques to predict the application behavior without requiring re-execution of application components. Experimental results obtained clearly show the effectiveness of our proposed approach with high prediction accuracy.