Order@Cloud: A VM organisation framework based on multi-objectives placement ranking

This paper presents the implementation and tests of a flexible and extensible framework, named Order@Cloud, that improves the Virtual Machine placements of a Cloud. It receives new VMs on the Cloud and organises them by relocating their placements based on the Multiple-Objectives of the environment. These Objectives are represented by Rules, Qualifiers and Costs, which can be easily added, extended and prioritised. Based on Evolutionary and Greedy Searches, Order@Cloud theoretically guarantees the adoption of a better set of Placements. More specifically, it seeks the non-dominated solutions (Pareto Set) and compares them considering the implementation cost of the scenario and its benefits. In contrast to existing solutions, that address specific objectives, our framework was devised to be objective-agnostic and easily extensible, which enables the implementation of new and generic prioritised elements. To understand the applicability and performance of our solution we conducted experiments using a real Cloud environment and discuss its performance, flexibility and optimality.