Hadoop is an open-source framework that use the map reduce architecture to store and analyse large datasets concurrently. As both inside and outside attacks may jeopardise a network's security, keeping this data safe is of the utmost significance. Although Hadoop does have a number of different authentication methods, only it takes into consideration a subset of possible security solutions. For this reason, we propose a technique that evaluates the efficiency of the encrypted multi-node hadoop cluster. As far as can be told, data may be stored in a file system called Hadoop Distributed File System (HDFS) that is shared across a group of computers. Our technology allows for the RSA-SHA encryption of any file or directory tree, making it a data-centric security solution. A K-means clustering mapreduce technique is used for streamlined distributed processing of massive volumes of files on clusters. Overall, our technology strengthens data protection while also facilitating simple and rapid access to previously stored files.