期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2018
卷号：6
期号：3
页码：2061
DOI：10.15680/IJIRCCE.2017.0603051
出版社：S&S Publications
摘要：Data is been increasing at an exponential rate in recent times. The data collected has to be processed andanalyzed carefully. Traditional algorithms and technologies are insufficient to process this data. The big data evolutionis dominating in today’s market. Many companies are exploring opportunities in this area of big data. To solve thisproblem Big Data frameworks are required. Hadoop and Spark, which are used to efficiently process the large data.With multiple big data frameworks available on the market, choosing the right one is a challenge. A detailedcomparison is done between the two techniques based on their performance in terms of memory, execution time, speedand applications used. Experimental results show that Spark is more efficient than Hadoop. However, spark requireshigher memory allocation, so the choice depends on performance level and memory constraints.
