An Osmosis-Based Intelligent Agent Scheduling Framework for Cloud Bursting in a Hybrid Cloud

Background

Hybrid cloud scheduling is a multi-objective optimization problem. The objectives are often conflicting. To satisfy multiple objectives, several heuristics in the literature focus on deciding which tasks to offload to a public cloud while minimizing execution times, renting costs, maximizing resource utilization and profit, minimizing energy consumption and satisfying deadlines. (Chittaranjan, Padmanabh, and Saxena, 2017; Singh and Chana, 2016; Tang et al., 2018; Zhang et al., 2016). To meet multiple objectives, the focus of several heuristics in literature is to decide which tasks to offload to a public cloud and in doing so how the execution time can be reduced and the cost kept at a minimum.

The Forward-Backward (FB) and subsequently the Forward-Backward-Refinement (FBR) Algorithm is modeled with the objective of minimizing cost and communications (Charrada & Tata, 2016). FPTAS (Fully Polynomial-time approximation scheme) (Reza, Farahabady, Lee, & Zomaya, 2014) uses Pareto-optimality in dealing with conflicting objectives of cost and performance and reaches optimal solutions with reasonable computational overhead.