Cyber Times International Journal of Technology & Management

Cloud computing is the most prominent technology in IT industry. With the advent of cloud computing it became easy to deploy virtual world which were earlier provided at a higher cost. However the resources have to be managed efficiently in order to gain more profit, thus the scheduler concept came into existence. Scheduler acts as a controlling agent which manages the virtual machines according to the physical properties of the infrastructure. Task allocation and resource management are the two main functionalities of the scheduler. The proposed scheduler is based on time based scheduling with the main objective of utilizing the idle resource leading to the maximum utilization of the physical resources and with minimum overlap in the task defined by the user. As per the implementation the prototype has been deployed on OpenStack Grizzly. Our proposed ‘Time-based Scheduler’ is built on OpenStack Grizzly, OpenStack Grizzly already has a scheduler called Nova scheduler, but the official document in OpenStack site states that more development has to be made for making it efficient .We propose fixed start and finish time to be given as input to the scheduler also an algorithm has been proposed which will assign the virtual machines to the physical resources. High quality scheduler depends on CPU utilization rate, throughput, turnaround time waiting time and response time proposed scheduler is being designed taking into account of these objectives. This paper aims to describe the ‘Time-based Scheduler’ and it describes how it enhances the scheduler of OpenStack.