首页 | 本学科首页   官方微博 | 高级检索  
     

支持构件迁移的分布式系统容错算法
引用本文:王勇,王忠群,刘涛,吴小兰. 支持构件迁移的分布式系统容错算法[J]. 计算机工程与设计, 2007, 28(15): 3566-3568,3730
作者姓名:王勇  王忠群  刘涛  吴小兰
作者单位:安徽工程科技学院,计算机科学与工程系,安徽,芜湖,241000;安徽工程科技学院,计算机科学与工程系,安徽,芜湖,241000;安徽工程科技学院,计算机科学与工程系,安徽,芜湖,241000;安徽工程科技学院,计算机科学与工程系,安徽,芜湖,241000
基金项目:安徽省自然科学基金 , 安徽省教育厅自然科学基金
摘    要:随着分布式应用技术的发展,构件的静态性限制了分布式系统在运行期间适应其可用资源变化的要求.为了动态适应系统可用资源,提高系统性能,支持构件可迁移的分布式应用研究受到广泛关注.针对提高构件可迁移分布式应用的可靠性,提出一种分布式系统容错算法.在准同步检查点算法的基础上,通过引入shadow-chain和通讯录机制,实现了构件间通信的可靠性和构件重定位的透明性,从而使得系统能够在较低代价下设置检查点,并能将出错后的系统回卷到一致性状态.

关 键 词:分布式系统  可迁移构件  检查点  回卷恢复  容错
文章编号:1000-7024(2007)15-3566-03
修稿时间:2006-08-15

Fault-tolerance algorithm for distributed systems supporting component migrating
WANG Yong,WANG Zhong-qun,LIU Tao,WU Xiao-lan. Fault-tolerance algorithm for distributed systems supporting component migrating[J]. Computer Engineering and Design, 2007, 28(15): 3566-3568,3730
Authors:WANG Yong  WANG Zhong-qun  LIU Tao  WU Xiao-lan
Affiliation:Department of Computer Science and Engineering, Anhui University of Technology and Science, Wuhu 241000, China
Abstract:With the development of distributed application technology,the static distributed component limits available resources to meet changing requirements.In order to enhance the system performance,distributed systems supporting component migrating has been wide concerned.To improve the dependability of distributed systems supporting component migrating,a distributed checkpoints algorithm is proposed.On the basis of index-based quasi-synchronous checkpoint algorithm,the reliability of communication between components and transparency of relocation of component is achieved by means of shadow-chain and address book mechanisms,this algorithm can take checkpoints at the lower price,and roll back the latest state of consistency.
Keywords:distributed systems  migratory component  checkpoints  rollback  fault-tolerance
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号