HBDCWS: heuristic-based budget and deadline constrained workflow scheduling approach for heterogeneous clouds

The predilection of scientific applications toward a high-performance computing system is attained through the emergence of the cloud. Large-scale scientific applications can be modeled as workflows and are scheduled on the cloud. However, such scheduling becomes even more onerous due to the dynamic and heterogeneous nature of cloud and therefore considered as a problem of NP-Complete. The scheduling of workflows is always constrained to QoS parameters. Most of the applications are bound to time and cost, which is observed to be the most crucial parameter. Therefore, in this paper, a heuristic-based budget and deadline constrained workflow scheduling algorithm (HBDCWS) has been proposed to utilize those applications that have the budget and deadline constraints. The novelty of the proposed work is to provide a simple budget and deadline distribution strategy where budget and deadline of workflow are converted to level budget and level deadline. Additionally, the level budget is again transferred to each task. This strategy not only satisfies the given constraints but also proves to be efficient for minimizing the makespan and reducing the cost of execution. Experimental results on several workflows demonstrate that the proposed HBDCWS algorithm finds a feasible solution that accomplishes the given constraints with a higher success rate in most cases.