Fecha de publicación: 30/10/2020 Fuente: Wipo "BigData" The invention discloses a data security auditing method and system based on a big data computing technology, and belongs to the field of big data security, and the method comprises the steps: collecting log data of a server, and transmitting the log data to a stream processing platform; receiving one or more log data, analyzing the log data, and sending the analyzed log data to at least one data destination; classifying the analyzed log data, judging whether the log data is real-time data or non-real-time data, and sending the real-time data to a stream processing platform for storage; sendingthe non-real-time data to a data center for storage; respectively analyzing and processing the log data to obtain an analysis result; and generating corresponding alarm information according to an analysis result and outputting the alarm information. The method has the advantages that log collection and storage are achieved based on flume, task scheduling and task monitoring are introduced, and flume log collection sources and flume log output sources are enriched; data security audit, alarm monitoring management and processing and security risk identification are realized based on the flink.