Components in the Pipeline

IEEE Software(2011)

引用 21|浏览0
暂无评分
摘要
Scientists commonly describe their data-processing systems metaphorically as software pipelines. These pipelines input one or more data sources and apply steps to transform the data and create useful results. Although conceptually simple, pipelines often adopt complex topologies and must meet stringent quality-of-service requirements that stress the software infrastructure used to construct the pipeline. The Middleware for Data-Intensive Computing (MeDICi) Integration Framework (MIF) is a component-based framework for constructing complex software pipelines. It supports composing pipelines from distributed heterogeneous software components and provides mechanisms for controlling the quality of service to meet demanding performance, reliability, and communication requirements.
更多
查看译文
关键词
middleware,pipeline processing,software quality,software reliability,MeDICi integration framework,data-processing pipelines,massive datasets,middleware for data-intensive computing,pipe-and-filter architecture,qualities of service,reliability,state-of-the-art scientific instruments,components,pipelines,scientific software,software engineering
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要