In this paper, we present a software framework designed to set up data flow chains in clusters of PCs or workstations. Communication between the components of such a chain is based on the publishersubscriber- paradigm including bridge components being used to transparently connect components on different nodes. In addition to these bridge components, the framework provides a number of further building blocks to shape the data flow, as well as templates for data sources, sinks, and analysis components. In the bridge components, a class library is used that abstracts from the actual network technology and protocol used and it already exists in versions for TCP/IP and SCI. We will give an overview of the frameworks architecture and present some results obtained in tests of the system.
CITATION STYLE
Steinbeck, T. M., Lindenstruth, V., Röhrich, D., Vestbo, A. S., & Wiebalck, A. (2002). A framework for building distributed data flow chains in clusters. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2367, pp. 454–464). Springer Verlag. https://doi.org/10.1007/3-540-48051-x_45
Mendeley helps you to discover research relevant for your work.