Information now streams from everyday life from telephones and charge cards and TVs and PCs; from the framework of urban areas from sensor-prepared structures, trains, transports, planes, extensions, and production lines. The information stream so quick that the aggregate gathering of the previous two years is currently a zettabyte. This colossal volume of information is known as large information. Huge Data alludes to advancements and activities that include information that is excessively various, quick-changing or gigantic for traditional advances, aptitudes, and framework to address productively. Said in an unexpected way, the volume, speed or assortment of information is excessively extraordinary. The volume of information with the speed it is created makes it troublesome for the present processing foundation to deal with enormous information. To conquer this downside, enormous information handling can be performed through a programming worldview known as MapReduce. Commonplace, execution of the MapReduce worldview requires arranged connected stockpiling and parallel preparing. Hadoop and HDFS by Apache are generally utilized for putting away and overseeing huge information. In this exploration paper, the creators recommend different strategies for taking into account the issues close by through MapReduce structure over HDFS. MapReduce strategy has been learned at in this paper which is required for actualizing Big Information examination utilizing HDFS and minimize time.