A Survey On Meliorate Data Distribution in Big data

The data distribution techniques hold an important role by having an influence on increasing the rate by which data can be processed as well as transferred in a cloud. A study has been made on various data distribution approaches one can look up on either in single cloud or multi cloud processing or storage systems. Centralized data distribution approach is the easiest way to set up in a client- server model but falls short of being an optimum model by the virtue of single point failure. A slight improvement over centralized is semi centralized approach. Peer to Peer approach stands firm by overcoming the single point failure, decreasing the latency in the data distribution. In this paper, various data distribution techniques along with P2P have been discussed in detail to uplift the data distribution in big data. A holistic study has been done on the hadoop ecosystem and the impact on load balancing which would eventually be taken as a parameter for testing its efficacy. Through all the required survey on various interlinked topics of data distribution mechanisms, we have inferred that P2P is still not an utopian scheme for data distribution and more work has to be contributed in coming days to remove all or few of its effecting demerits on data distribution.