摘要
本文基于工作站网络NOW(networksofworkstations)的特性,提出一个自适应可伸缩的进程迁移和协同凋度模型DASIC(dynamicadaptivescalableprocessmigrationandcoscheduling),试图解决其动态自适应要求越来越高的问题.DASIC模型在系统协同调度的基础上增加动态自适应的可伸缩功能,使其能够动态适应NOW环境下的变化,进行可伸缩的调度和负载平衡凋节.在保证工作站用户独占特性的同时,提高了整个系统资源的利用率,为当前NOW环境中系统资源充分利用的研究提供了有效的模型.
Based on the scalability on NOWs (networks of workstations), this paper proposes an adaptive migration and coscheduling model DASIC(dynamic adaptive scalable process migration and coscheduling). DASIC takes not only coscheduling to schedule in teractive processes simultaneously, but also adaptive migrating to satisfy the system's dy namics and scalability. It can grow, shrink and load balance to adjust the processes scheduling status in the process space, and the processor pool can be expanded or shrinked.
出处
《软件学报》
EI
CSCD
北大核心
1997年第10期752-759,共8页
Journal of Software
基金
国家863高科技项目
国家基础研究攀登计划资助
国家教委博士点基金
关键词
协同调度
负载平衡
迁移
操作系统
Adaptive, coscheduling, load balancing, migration, scalable