Monitoring of Cloud services is vital for both service providing organizations and consumers. The service providers need to maintain the quality of service to comply their services with the QoS parameters defined in SLA's such as response time, throughput, delay through continuous monitoring of services. The dynamic monitoring involves prediction of SLA violations and subsequent adaptation of the service compositions. The task of adaptation is in fact the task of discovering another plausible composition in the face of services recorded to have generated QoS violations. QoS- Driven Utility based service composition approach considers the individual user's priorities for QoS parameters and determines the overall utility measure of the service composition for the end user. In this work we present the problem of service composition adaptation as a multi-objective assignment optimization problem, which in turn is a NP-hard problem. The evolutionary algorithm GA with Tabu has been formulated as a Memetic and Pareto optimal approach for the adaptation problem and analyzed for efficiency in solving the problem.

Cloud computing is an emerging internet-based paradigm of rendering services on pay- as -per -use basis. Increasing growth of cloud service providers and services creates the need to provide a tool for retrieval of the high-quality optimal cloud services composition with relevance to the user priorities. Quality of Service rank-ings provides valuable information for making optimal cloud service selection from a set of functionally equiva-lent service candidates. To obtain weighted user-centric Quality of Service Composition, real-world invocations on the service candidates are usually required. To avoid the time-consuming and expensive real-world service invocations, this paper proposes framework for predic-tion of optimal composition of services requested by the user. Taking advantage of the past service usage experi-ences of the consumers more cost effective results are achieved. Our proposed framework enables the end user to determine the optimal service composition based on the input weight for individual service Quality of Service. The Genetic algorithm and basic Tabu search is applied for the user-centric Quality of Service ranking prediction and the optimal service composition. The experimental results proves that our approaches outperform other competing approaches.