MapReduce with Components for Processing Big Graphs

  • Cenez Araújo de Rezende UFC
  • Francisco Heron de Carvalho Junior UFC

Abstract


BigGraph applications have been mostly implemented with largescale parallel processing frameworks based on MapReduce. However, it continues to be challenging to meet their particular requirements, even despite the emerging of alternative models, such as Pregel. This paper introduces a component-oriented design for graph processing frameworks, by using HPC Shelf, a component-based cloud computing platform for HPC services.
Keywords: Parallel processing, Computational modeling, Connectors, Feeds, Cloud computing, Synchronization, Contracts, High Performance Computing, Cloud Computing, Component-Based Software Engineering, Graph Processing
Published
2018-10-01
How to Cite

Select a Format
DE REZENDE, Cenez Araújo; DE CARVALHO JUNIOR, Francisco Heron. MapReduce with Components for Processing Big Graphs. In: BRAZILIAN SYMPOSIUM ON HIGH PERFORMANCE COMPUTING SYSTEMS (SSCAD), 19. , 2018, São Paulo. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2018 . p. 108-115.