A penalty-based grouping genetic algorithm for multiple composite SaaS components clustering in Cloud

摘要: Software as a Service (SaaS) in Cloud is getting more and significant among software users providers recently. A SaaS that delivered composite application has many benefits including reduced delivery costs, flexible offers of the functions decreased subscription cost for users. However, this approach introduced new problem managing resources allocated to SaaS. The resource allocation been done at initial stage may be overloaded or wasted due dynamic environment Cloud. typical data center management usually triggers placement reconfiguration order maintain its performance well minimize used. Existing approaches often ignore underlying dependencies between components. In addition, also comply with constraints terms requirements, requirement SLA. To tackle problem, paper proposes penalty-based Grouping Genetic Algorithm multiple components clustering main objective used by component without violating any constraint. Experimental results demonstrate feasibility scalability proposed algorithm.