SLA-based optimization of energy and migration cost on openflow platform

Platform as a Service (PaaS) is one of the cloud computing services for offering virtual operating systems to clients. The rave over cloud computing services is due to its obvious advantages compared to traditional network services. Those advantages include: high computational power, high scalability, and lower requirements needed from end devices. However, one of the challenges for cloud service providers is related to Service Level Agreements (SLAs), which specify constraints on the provided Quality of Service. The profit of the cloud provider depends on how they meet the SLA. Generally, the greater the size of the datacenter, the better performance clients can achieve. Moreover, the energy cost related to the size of the datacenter results in basic trade-offs with client satisfaction. In this thesis, a dynamic and migration algorithm is proposed to minimize the cost of energy in consideration of SLAs. An emerging Software Defined Networks (SDN) technology called OpenFlow technology is utilized in our local testbed. Simulation and experimental results demonstrate that the efficiency of the resource usage and reduced power consumption of the cloud can coexist with SLAs while keeping the cost of penalties and power consumption to minimum