In this paper present survey on Data mining, Data mining using Rough set theory and Data Mining using parallel method for rough set Approximation with MapReduce Technique. With the development of information technology data growing at an incredible rate, so big data mining and knowledge discovery become a new challenge. Big data is the term for a collection of data sets which are large and complex, it contain structured and unstructured both type of data. Data comes from everywhere,  posts to social media sites, digital pictures and videos etc this data is known as big data. Useful data can be extracted from this big data with the help of data mining. Data mining is a technique for discovering interesting patterns as well as descriptive, understandable models from large scale data. Rough set theory has been successfully applied in data mining by using MapReduce programming technique. We use the Hadoop MapReduce System as an Implementation platform. The lower and upper approximations are two basic concept of rough set theory. A parallel method is used for the effective computation of approximation and is improving the performance of data mining. With the benefits of MapReduce  The MapReduce technique, received more attention from scientific community as well as industry for its applicability in big data analysis it makes our approach more ideal for executing large scale data using parallel method .In this paper we have presented working and execution flow of the MapReduce Programming paradigm with Map and Reduce function. In this work we have also briefly discussed different issues and challenges that are faced by MapReduce while handling the big data. And lastly we have presented some advantages of the Mapreduce Programming model.