Effective Big Data Analysis Based on Sockets. Application to Biomedical Data Processing

This paper implements a cluster system based on personal computers with the involvement of traditional TCP and sockets based on the virtual interface architecture with a mechanism of data splitting and the possibility of processing homogeneity. We investigate the productivity of such system, and presented experimental results that prove significant improvements in performance due to the reorganization of individual application components, which stimulates the growth of their scalability.