期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2019
卷号：19
期号：5
页码：113-117
出版社：International Journal of Computer Science and Network Security
摘要：In today's world data is becoming gold. And with the passage of time it has started to grow exponentially and being referred as Big Data. Big Data includes data sets that cannot be handled by common tools because of its complexity in terms of volume, velocity and variety. One of the challenges of big data processing is to process the data within given time and memory. The papers presents a concept of optimization of big data processing using software define networking. For this, the two well-known big data frameworks Hadoop and Spark are deployed over SDN. The communication behavior and pattern among the nodes of Hadoop and spark were studied and optimized by varying the maximum transmission unit (MTU) of network packet. After executing many standard big data benchmark d including HiBench Benchmark Suite, it is clearly depicted that how big data frameworks performs better in any SDN environment in terms of reduced processing time with the increase of MTU.
关键词：Hadoop; Apache Spark; Big Data ; SDN; OpenFlow; Benchmarks
