An efficient framework to handle integrated VM workloads in heterogeneous cloud infrastructure

When a virtual machine migrates to a host, virtual machine placement is used to estimate the suitability of available host. The virtual machine placement algorithm evaluates the performance data for the host, settings of hardware, resource requirements, and host ratings to provide the best placement on host. Generally, the service providers cannot manage or control the cloud infrastructure including operating system, servers, storage and the placement of service components. This lack of consequence is a deterrent to cloud adoption for some customers. This paper presents an approach for customers to handle the integrated virtual machine workloads. The fuzzy decision making approach, integer linear programming and cost factors such as strategic decision, selection of cloud computing services and its types are considered in this paper. The fuzzy quantifiers are used for selecting the number of virtual machines in the environment and it optimizes the demand of the cloud resources in distributed environment. Integer linear programming constraints have been used to enable infrastructure providers to optimize the placement of virtual machines and algorithms to provide service provider effect. The experimental results show that the proposed approach provides better performance than all other existing techniques such as Markov chain model and optimal crossover genetic algorithm.