Abstract: | With the rapid development of the Internet industry, the data scale grows bigger and bigger. At the same time, big data also shows distinct streaming characteristics. Traditional batch data processing mode based on MapReduce becomes unable to meet the demand of streaming data processing. Therefore, the more efficient specialized stream computing systems is used more and more widely. Firstly, this paper introduces the basic concepts and characteristics of stream big data, compares several typical streaming data processing solutions, and analyzes their system architecture, performance and application scenarios. Finally, the future work on big data processing technology is discussed. |