Survey on data failure handling methods of streaming data

The main purpose of this paper is to provide an exhaustive analysis of different techniques available for handling the data failures in big data analytics. This article surveys different failure handling methods available for streaming data and assesses the advantages and drawbacks of each of these methods based on various performance metrics such as scalability, consistency, speed of access, real-time processing, data size etc. Moreover to the techniques, Tools and description of the frameworks used is also discussed along with their strengths and challenges. Handling the failures of streaming data is a challenging task. Recently Apache spark stream guarantees the zero data loss in fault tolerance.