Suboptimal mechanism for load balancing in cloud environment

Abstract

Load balancing is important and major issue for cloud networks due to its elastic nature, user's unpredictable behavior and difficulty faced to foresee the request issued to the server. In order to efficiently utilize cloud resources, we are presenting a new methodology for dynamic resource allocation based on the suboptimal mechanism of aggregation i.e. GIT (greedy incremental tree). The technique improves path sharing at time of migration during load balance process. A cloud workload network is constructed based on-Greedy heuristics having the characteristics to build incrementally diverse set, selecting one compute node at a time to maximize diversity measure. The virtual machine manager can rely on the rules and can make efficient decision for optimal provisioning of virtual machine based on the aggregated route.