摘要
提出一种基于树形计算网格的自适应调度模型,实现对小粒度独立任务和用户大作业的自适应最优调度.通过对网格环境的实时检测,给出了基于节点负载状况、节点任务执行时间和任务特性的自适应调度算法.实验证明该任务调度模型在负载平衡和容错方面具有良好的性能.
This paper presents an adaptive scheduling model based on tree grid, This model can schedule fine - grained tasks and big jobs adaptively. Through the Real - time detection of grid, an adaptive scheduling algorithm is proposed which is based on load state, task execution time and task properts. The experiments show that the proposed algorithms for the scheduling problem have good performance on load - balancing and fault tolerant.
出处
《微电子学与计算机》
CSCD
北大核心
2008年第4期56-58,62,共4页
Microelectronics & Computer
基金
河南省科技攻关项目(0524220042)
关键词
网格计算
自适应调度
任务池
计算池
grid computing
adaptive scheduling
task pool
computing pool