Cloud Computing has managed to attract the entire buzz in the growing era of technology due to its on-demand services for resource request. Despite of the enormous growth of cloud computing, there are many problems related to resource allocation in cloud that are still unaddressed. Current work for resource allocation strategy focuses on various methods to place Virtual Machine per appropriate requests. The current state of art focuses on the dynamic nature of the work load on cloud. But there is still scope of improvement in the resource allocation strategies that have been proposed in terms of well-balanced network even at the resource contention.
This study proposes a hybrid model composed of lease methodology and dynamic load balancing algorithm, with an attempt to overcome the problems of resource contention and starvation and a well-balanced network even at the input of varying loads. An attempt to increase the CPU utilization and throughput along with no request rejection is taken. The work also retains the lease options for its clients thus maintaining the anti-starvation for pre-emptible requests.