Qos oriented dynamic and energy Aware resource allocation schemes For high performance computing Clouds

Abstract

In the advanced field of computers and networking, Cloud computing becomes an imperative and domineering topic. Cloud computing is an Internet based computing where the dynamic and virtualized resources are given and distribution of these resources to various tasks are performed by the computers. It includes variety of computing like network storage technologies, load balance, distributed computing, parallel computing, utility computing and virtualization. The computing resources available in a cloud computing are assigned to the corresponding task demanded by the clients on the cloud using an effective dynamic resource allocation (DRA) system. The resources are optimally allocated by using a multi-objective OA (Optimization Algorithm) by taking the objective function as total cost, time consumption, and minimum CPU utilization. During allocation, many resources and incoming tasks are available. Hence matching the tasks to the suitable resources from the available resources is quite challenging. A task can get executed with all the resources but finding the suitable resource with less cost, time and utilization of CPU is the objective here. This objective is best achieved through optimization. In cloud computing, resource allocation is a vital task which uses the available resources and allocates them to the demands induced by the users by making use of optimal number of servers. The promising resource nodes for operating a task in cloud computing should be selected according to the attributes of the task with the possibility of attaining balance among the QoS requirement at reduced cost. Thus the resource allocation process becomes a multi-objective optimization problem which should establish a trade-off between the cost and the establishment of suitable service. In this work, the dynamic resource allocation (DRA) is performed by making use of improved flower pollination algorithm (Im-FPA). Flower pollination algorithm dynamically allocates resources to the tasks newline newline newline