Fig. 2From: A parallelization model for performance characterization of Spark Big Data jobs on Hadoop clustersAmdahl’s law with various percentages of serial workBack to article page