Task Allocation and Re-allocation for Big Data Applications in Cloud Computing Environments

Abstract

Resource allocation for Big data streams in cloud systems involves selecting the appropriate cloud resources. Since incorrect resource allocation results in either under provisioning or over provisioning, accurate resource allocation becomes challenging in Big data applications. Hence, the objective of this work is to design an optimal solution for resource allocation for minimizing the network bandwidth and response delay. In this paper, a task allocation and re-allocation mechanism for Big data applications is designed. It consists of two important agents: RE-allocation Agent (REA) and Resource Agent (RA). The RA is responsible for mapping the user requirements to the available VMs. The REA monitors the resources and chooses the VMs for resource reconfiguration. Then, it dispatches an allocation or de-allocation request to RA, running in the physical system, based on the varying requirements of virtual machines. Experimental results show that the proposed TARA has less execution time and achieves better utilization of resources, when compared to existing tool.