HHMFO-DVFS for An Optimal Makespan-Energy-aware Workflow Scheduling in Cloud

Abstract Nowadays, various workflow applications have expanded that need to be stored and processed on powerful and heterogeneous computing platforms such as cloud environments. However, this issue increases energy consumption in cloud servers, which raises expenses for cloud computing resource providers. By reducing Makespan time, Virtual Machines (VMs) spend less time performing workflows, resulting in lower system energy usage. Cloud systems must also meet deadlines and be reliable. So, the main objectives are to reduce the cloud computing system's energy consumption and the Makespan time, while maintaining reliability and deadline constraints. To achieve these, the Harris Hawks and Moth-Flame Optimizer (HHMFO) is utilized to pick the optimum VMs for process execution while optimizing the energy consumption and Makespan time parameters, as well as Maintaining the reliability and deadline constraints. As a consequence, each time the HHMFO selects multiple VMs, the workflow Makespan time is lowered on the selected VMs while keeping reliability and deadline. The static and dynamic energy consumption of processes is then lowered by utilizing Processor Merging (PrM) and Dynamic Voltage Frequency Scaling (DVFS) methods on the Actual Finish Time (AFT). The experimental findings on real workflows demonstrate that the suggested strategy is more successful than the comparison approaches.