Dynamic information stream handling utilizing constant programming model for processing large data sets with a parallel, distributed algorithm on a cluster is at present a high worry as the measure of information being created is expanding step by step with the development of the Internet of Things, Big Data and Cloud. Big data are portrayed by immense volume that can land with a high speed and in various organizations from numerous sources. Accordingly, continuous programming model strategies ought to be fit for preparing the information to separate an incentive out of it by tending to the issues identified with these qualities that are related to information streams. In this work, we asses and break down the ability to exist Map-Reduce and Spark procedures to deal with dynamic information streams and we introduce whether the current systems are important in the current circumstance.