R-FirstFit: A Reservation Based First Fit Priority Job Scheduling Strategy and Its Application for Rendering

When high performance computing platforms are widely used for rendering application, scheduling of rendering jobs becomes crucial due to the job starvation and resource fragmentations problems which consist in the existing scheduling strategies. Meanwhile, these strategies are rarely balance fairness and performance. Aiming at these issues, we propose a reservation based First Fit priority (R-First Fit) job scheduling strategy for cluster rendering system. In this strategy, the reservation based First Fit method is used to maximize the resources utilization for rendering system when the jobs are blocked. The proposed R-First Fit strategy is compared with other scheduling strategies (FCFS and First Fit) under different evaluation metrics. The experiment results demonstrate that the proposed strategy outperforms in many cases than the existing ones while guaranteeing the fairness of scheduling jobs.