摘要
针对目前工作流管理系统建立数据流存在的问题,引入流内数据流和流外数据流的概念,在分析数据流特点的基础上,提出了一种分布式环境下管理工作流数据的体系结构.在该结构中,运行层的运行节点负责控制流,数据层的数据节点负责数据流,运行节点通过执行接收活动来触发数据层的数据复制机制,进而在运行节点间建立所需的数据流.与现有的实现数据流的方法相比,本文提出的数据描述方法和数据访问机制不仅为流外数据流提供了数据通道,同样为流内数据流的描述和实现提供了方便,对设计工作流管理系统中的数据管理具有较高的应用价值.
In order to solve the problem in establishing data riot data flow and exterior data flow are introduced in this flow in existing workflow management systems, conceptions of intepaper. Based on analyzing characteristics of data flows, a new architecture of data management for workflow in distributed environments is put forward. By making that the running nodes in the running level are responsible for the control flow, the data nodes in the data level are responsible for the data flow, and the execution of receiver activities on the running nodes triggers the data replication mechanism of the data level, the required data flow between running nodes can be established. Compared with the existing methods of data flow implementation, the proposed method of data description and data access mechanism provide a data transfer means for the exterior data, and convenient the description and implementation for the interior data. The proposed method has value in designing data management in workflow management system.
出处
《小型微型计算机系统》
CSCD
北大核心
2009年第2期259-265,共7页
Journal of Chinese Computer Systems
基金
国防装备科研基金项目(EP060106)资助
关键词
数据流
控制流
工作流数据
数据描述
数据访问机制
data flow
control flow
workflow data
data description
data access mechanism
