An Evaluation of Job Scheduling Based on Distributed Energy Generation in Decentralized Data Centers

Compared to a centralized data center, decentralized data centers have an advantage of disaster recovery, but costs more expensive. In order to relax the increase of the cost, it is important to utilize the distributed energy resource in the decentralized data centers. Because power is necessary for processing jobs in data centers, it is necessary to make a job schedule that is adjusted to the distributed energy. In this paper, we apply the existing job scheduling algorithms that have been used for single data center to decentralized data centers. For the purpose of clarifying the research issue, we evaluate the deadline violations of jobs and distributed energy use when applying the scheduling algorithms.