Automated design of a parallel distributed system for streaming data processing

Abstract

The development of high-level algebra-algorithmic tools for automated parallel program design for a distributed dynamically scalable fault-tolerant computing platform is proposed. The usage of the developed tools is illustrated on an example of designing a parallel distributed system intended for processing a large amount of streaming data. As an example of data source, there was chosen the Twitter social network. The deployment of the distributed system to the cloud platform is considered and experiment results are given.