Job scheduling using Minimum Variation First algorithm in cloud computing

Nowadays, the problems are becoming more complicated due to the development of fields related to science and engineering. Cloud computing plays a major role to figure out these complicated problems. Cloud computing is generally categorized into computation intensive and storage intensive model. Cloud collects congregate myriad number of requests from the user (i.e. treated as batch jobs). Hence, scheduling algorithm plays a major role for effectively scheduling of the jobs to the underlying resources scattered in and around the universe. The resources are linked through high speed network. This paper posits Minimum Variation First algorithm (MVF) for effective scheduling of batch jobs. The difference between the expected execution time on the job and its corresponding deadline is recognized as a necessary parameter for allocating the resource for a job. The involvement of this paper is considered as twofold. First, the deadline based jobs are scheduled using the proposed MVF algorithm that will schedule with uniform and non-uniform deadline based jobs. Second, the jobs are scheduled using improved MVF (iMVF) algorithm for avoiding starvation. The experimental results show the performance of these algorithms (MVF and iMVF algorithm) is better compared to other algorithms using CloudSim.