基于优先级的任务调度与负载均衡模型研究

Study of Task Scheduling and Load Balancing Models Based on Priorities

在分布式计算环境下,为了有效地利用计算资源、快速完成协同计算任务,提出了基于优先级的任务调度与负载均衡模型.首先根据就绪任务队列和任务调度器所处的位置以及两者之间的关系,将任务调度划分为集中式任务调度和非集中式任务调度两种方式,在此基础上,利用时间Petri网建模技术,分别给出了采用这两种任务调度方式的、基于优先级的任务调度与负载均衡模型,并对各种模型的特点进行了详细分析.以此模型为基础,可以利用现有的时间Petri分析技术,对采用上述任务调度方式的任务调度算法进行模拟和分析,以便找出满足给定条件的最优的任务调度算法.

To utilize the computing resources effectively and accomplish the cooperative task rapidly in the distributed environment, the priorities based task scheduling and load balancing models are proposed. According to the positions of the ready task queues and task schedulers, the task scheduling is divided into two classes, i.e. , centralized task scheduling and decentralized task scheduling. Based on this result of classification, the priorities based task scheduling and load balancing models are proposed respectively by using time Petri net modeling techniques, and the features of these models are analyzed in detail. With the help of these models, the specific task scheduling algorithms using these task scheduling modes can be simulated and analyzed by using the existing time Petri net analyzing techniques, so as to find out the optimized task scheduling algorithm.