Task scheduling in real-time industrial scenarios

Abstract

Task scheduling for microservice-oriented industrial software is a complex process. It is a real-time process where multiple task attributes should be considered and different tasks should be processed in parallel. To address this problem, a scheduling model including a task model and resource model applicable to industrial scenarios is established. Then a task importance function for evaluating the importance of tasks is designed. After that, a Dynamic Importance-aware Online Scheduling Algorithm (DIOS) is proposed, which ranks and schedules all the tasks in decreasing order of importance. To further improve the efficiency and adaptability of the DIOS algorithm, resource reservation, preemptive scheduling and online adaptive tuning methods are established. Finally, simulation results in two typical industrial software scenarios show that the proposed algorithm performs significantly better than the widely used algorithms such as the First Come First Serve Algorithm (FCFS), the Earliest Deadline First Algorithm (EDF), the Least Laxity First Algorithm (LLF) and the Fixed Priority Scheduling Algorithm (FPS).