Hadoop performance analysis on Raspberry Pi for DNA sequence alignment

Abstract

The rapid development of electronic data has brought two major challenges, namely, how to store big data and how to process it. Two main problems in processing big data are the high cost and the computational power. Hadoop, one of the open source frameworks for processing big data, uses distributed computational model designed to be able to run on commodity hardware. The aim of this research is to analyze Hadoop cluster on Raspberry Pi as a commodity hardware for DNA sequence alignment. Six B Model Raspberry Pi and a Biodoop library were used in this research for DNA sequence alignment. The length of the DNA used in this research is between 5,639 bp and 13,271 bp. The results showed that the Hadoop cluster was running on the Raspberry Pi with the average usage of processor 73.08%, 334.69 MB of memory and 19.89 minutes of job time completion. The distribution of Hadoop data file blocks was found to reduce processor usage as much as 24.14% and memory usage as much as 8.49%. However, this increased job processing time as much as 31.53%.