分布式任务调度的智能化及其实验系统KZ2/FRD

INTELLIGENT SCHEDULING OF DISTRIBUTED TASKS AND ITS EXPERIMENTAL SYSTEM KZ2/FRD

在一个大型分布式系统中,传统的调度方法存在着知识的不确定性、不完整性、调度不稳定性和策略单一性等问题.为了解决这些问题,本文提出一个智能分布式任务调度的新方法,并介绍了实验系统KZ2/FRD的实现和性能分析.

In a large distributed system, tranditional scheduling algorithms possess the problems such as uncertainty and incompleteness of knowledge, instability of schedule, unistra-tegy, etc. This paper proposes an idea of intelligent scheduling of distributed tasks to solve these problems. An experimental system KZ2/FRD and its performance are discussed.