Cloud Scheduling Heuristic Approaches for Load Balancing in Cloud Computing

Cloud computing provides a framework for remote users to access diverse resources in DC (Data Centers) to compute High-Performance Computing workloads. In the cloud, physical resources are virtualized to provide user services via VMs (Virtual Machines). Task scheduling is a crucial component of the cloud, and successful VM usage by CSPs (Cloud Service Providers) necessitates an optimum task-scheduling algorithm. An ideal scheduling heuristic must be efficient, fair, & starvation-free to provide a shorter timeline with better resource use. However, static heuristics frequently result in inefficient and ineffective resource use on the Cloud. The fundamental issue in the CC environment is task scheduling, which is critical in optimizing overall execution time. This paper compares scheduling heuristic algorithms such as DRALBA, RALBA, DLBA, Min-min, Max-min & Round Robin depending upon makespan, throughput, and ARUR using workflows on synthetic workload and Google a realistic workload called GoCJ datasets. According to this study, the existing DRALBA approach outperforms in terms of performance parameters than the other approaches over both synthetic and GoCJ data.