Service Level Agreement Driven Adaptive Resource Management For Web Applications on Heterogeneous Compute Clouds

Cloud computing is an emerging topic in the field of parallel and distributed computing. Many IT giants such as IBM, Sun, Amazon, Google, and Microsoft are promoting and offering various storage and compute clouds. Clouds provide services such as high performance computing, storage, and application hosting. Cloud providers are expected to ensure Quality of Service (QoS) through a Service Level Agreement (SLA) between the provider and the consumer. In this research, I develop a heterogeneous testbed compute cloud and investigate adaptive management of resources for Web applications to satisfy a SLA that enforces specific response time requirements. I develop a system on top of EUCALYTPUS framework that actively monitors the response time of the compute resources assign to a Web application and dynamically allocates the resources required by the application to satisfy the specific response time requirements