A QoS-Enabled Load Balancing Approach for Cloud Computing Environment Join Minimum Loaded Queue (JMLQ)

Abstract

Cloud computing delivers the on-demand virtualized resources to its consumer for servicing their request on a metered basis. During the high demand of cloud resources the load on system increases that may unbalance the system which affects the quality of service parameters (QoS) adversely that leads to violations of service level agreement (SLA). Role of load balancing is significant in such an environment as it enhances the distribution of workload across multiple devices for example across network links, a cluster of servers, disk drives, etc. The present research work introduced a multi scheduler for balancing the load across the system that aims to optimize the QoS parameters such as response time, resource utilization, and the average waiting time by exploiting these virtual resources in the cloud environment. The performance of the proposed approach analyzed and tested in CloudSim that to optimize these parameters for the current approach. The authors found that our QoS enabled JMLQ approach achieved better results in comparison to our previous JMLQ approach and other variants.

A distributed quality of service-enabled load balancing approach for cloud environment

“…The incoming task probes the first dispatcher I-queue with ID 0 having identification D 0 to get a minimum loaded VM; If it finds VMs within I-queue, the task is mapped to the first minimum loaded VM, and it is removed from the corresponding I-queue. On the other side, if the attached I-queue is empty, the task is directed toward the adjacent dispatcher I-queue to get a minimum loaded VM. VM mapping to I-queues: EQoS-enabled JMLQ uses the d-limit concept of our previous proposed approach JMLQ (Sharma et al , 2019c, 2021) to set an upper bound for the number of VMs to be placed in an I-queue. After completion of the task, the VM is directed toward the beginning dispatcher’s I-queue for joining it.…”

“…After completion of the task, the VM is directed toward the beginning dispatcher’s I-queue for joining it. Here, the dispatcher’s I-queue length is determined to check the d-limit, the VM joins the I-queue if the length is less than the d-limit otherwise drifted toward the adjacent dispatcher I-queue and so on (Sharma et al , 2019a). The VMs are evenly distributed among I-queues based on d-limit at moderate load using dispatcher identity.…”

“…

Purpose -During high demand for the virtualized resources in cloud environment, efficient task scheduling achieves the desired performance criteria by balancing the load in the system.Design/methodology/approach -It is a task scheduling approach used for load balancing in cloud environment. Task scheduling in such an environment is used for the task execution on a suitable resource by considering some parameters and constraints to achieve performance.Findings -The presented mechanism is an extension of the previous proposed work quality of service (QoS)-enabled join minimum loaded queue (JMLQ) (Sharma et al, 2019c). The proposed approach has been tested in the CloudSim simulator, and the results show that the proposed approach achieves better results in comparison to QoS-enabled JMLQ and its other variants in the cloud environment.

…”

“…Findings -The presented mechanism is an extension of the previous proposed work quality of service (QoS)-enabled join minimum loaded queue (JMLQ) (Sharma et al, 2019c). The proposed approach has been tested in the CloudSim simulator, and the results show that the proposed approach achieves better results in comparison to QoS-enabled JMLQ and its other variants in the cloud environment.…”

“…Join minimum loaded queue (JMLQ) allocates the task to the virtual machine (VM) by avoiding such overheads. The system follows a mechanism in which a task is allocated to a minimum loaded VM by avoiding extra communication overheads (Sharma et al, 2020b(Sharma et al, , 2019a.…”