Big Data Solution: Improvised Distributions Framework of Hadoop

摘要

Big data technology (synonymous with Apache Hadoop) is growing rapidly due to its low cost, scalable data processing capabilities, and high fault tolerance. After the release of Hadoop 1.0, the organizations started using this technology but faced some difficulties. At this stage, Hadoop Framework was premature in fulfilling the expectations of different organizations, which looked for an alternative of Apache Hadoop. Therefore, Hadoop vendors improvised the open source core Apache Hadoop release and made it enterprise ready. Hadoop Distributions presented a unified product: an added apache repository, new tools, security and central administration all together so that the organizations did not have to spend time assembling all these essentials into a single functional piece. The most accepted Hadoop Distributions existing in the market today are Cloudera CDH, Hortonworks HDP, and MapR M5. This paper discusses different features of these Hadoop vendor distributions and makes comparisons based on the functionality and contribution of each vendor in the big data market.