A workload clustering based resource provisioning mechanism using Biogeography based optimization technique in the cloud based systems

Cloud computing is one of the rapidly growing distributed computing technologies, and cloud-based applications have increased significantly in recent years. The amount of cloud resources and the number of cloud user are important metrics that affect the management of the cloud-based applications. Since the volume of traffic to cloud-based applications grows, the resource provisioning as one of challenging issues to serve time-varying and heterogeneous workloads in resource management scope to be considered. In this paper, we propose a workload clustering-based resource provisioning mechanism for executing cloud-based applications with heterogeneous workloads. Our proposed mechanism utilized biogeography-based optimization (BBO) technique with K-means clustering to classify the cloud workloads according to their quality of service (QoS) requirements. Besides, we used Bayesian learning technique to specify suitable resource provisioning actions to satisfy the QoS requirements of cloud-based applications. The simulation results obtained through simulation demonstrate that the proposed solution reduces the delay, SLA violation ratio, cost, and energy consumption compared with workload clustering-based resource provisioning mechanisms.