Improving fault-tolerance by replicating agents
Improving fault-tolerance by replicating agents
复制标题
通过复制代理提高容错能力
DOI：
--
复制
发表时间：
2002
期刊：
影响因子：
--
通讯作者：
R. Deters
中科院分区：
文献类型：
--
作者：
Alan Fedoruk;R. Deters
尽管在开发多代理系统上花费了大量的精力，但实际部署的系统数量却少得惊人。已开发系统和已部署系统之间存在巨大差距的原因之一是它们的脆弱性。由于缺乏集中控制组件，很难检测和处理单个代理的故障，因此有可能出现严重影响系统性能的故障传播。通过在多代理系统中复制单个代理来使用冗余是提高容错性的一种可能方法。不幸的是，复制的引入会增加复杂性和系统负载。在本文中，我们研究了透明代理复制的使用，这是一种代理复制出现并作为一个实体的技术，从而避免了系统复杂性的增加并最大限度地减少了额外的系统负载。本文定义了透明代理复制，并确定了使用它的主要挑战。特别关注代理间通信、读写一致性、资源锁定、资源综合和状态同步。介绍了FIPA-OS框架的透明代理复制的实现，并给出了在实际多代理系统中对其进行测试的结果。
Despite the considerable efforts spent on developing multi-agent systems the actual number of deployed systems is surprisingly small. One of the reasons for the significant gap between developed and deployed systems is their brittleness.The absence of centralized control components makes it difficult to detect and treat failures of individual agents thus risking fault-propagation that can seriously impact the performance of the system. Using redundancy by replication of individual agents within a multi-agent system is one possible approach for improving fault-tolerance. Unfortunately the introduction of replicates leads to increased complexity and system load. In this paper we examine the use of transparent agent replication, a technique in which the replicates of agents appear and act as one entity thus avoiding an increase in system complexity and minimizing additional system loads. The paper defines transparent agent replication and identifies the key challenges in using it. Special attention is given to the inter-agent communication, read/write consistency, resource locking, resource synthesis and state synchronization. An implementation of the transparent agent replication for the FIPA-OS framework is presented and the results of testing it within a real-world multi-agent system are shown.