Enormous Data has long been the subject of interest for Computer Science aficionados around the globe, and has\nincreased much more conspicuousness in the later times with the constant blast of information coming about\nbecause of any semblance of online networking and the journey for tech monsters to get entrance to more profound\ninvestigation of their information. MapReduce and its variations have been very fruitful in actualizing vast scale\ninformation concentrated applications on ware groups. Then again, a large portion of these frameworks are\nmanufactured around a non-cyclic information stream demonstrate that is not suitable for other famous applications.\nOriginal MapReduce executes jobs in a simple but rigid structure fashion. MapReduce transforms step (“map”), a\nsynchronization step (“shuffle”), and a step to combine results from all the nodes in a cluster (“reduce”).\nAccordingly to defeat the inflexible structure of guide and diminish we proposed the as of late presented Apache\nSpark – both of which give a handling model to breaking down enormous information. The main contender for\n\"successor to MapReduce\" today is Apache Spark. Like MapReduce, it is a broadly useful motor, however it is\nintended to run numerous more workloads, and to do as such much speedier than the more seasoned framework. In\nthis paper we contrast these two systems along and giving the execution examination utilizing a standard machine\nconsidering so as learning calculation for bunching (K-Means) and through considering some different parameters\nlike scheduling delay, speed up, energy consumption than the existing systems.