Evaluating MapReduce for profiling application traffic

摘要: The use of MapReduce for distributed data processing has been growing and achieving benefits with its application different workloads. can be used traffic analysis, although network traces present characteristics which are not similar to the type commonly processed through MapReduce. Motivated by profiling due lack evaluation analysis peculiarity this kind data, paper evaluates performance in packet level DPI, analysing scalability, speed-up, behavior phases. experiments provide evidences predominant phases job, show impact input size, block size number nodes, on completion time scalability.