A big data storage technology based on domestic autonomous controllable characteristics

In view of the characteristics of the target product, such as diverse feature data types, huge capacity, and large number of unstructured and semi-structured data, based on the domestic autonomous controllable platform, we adopt the hybrid storage architecture of massive feature data files combining relational database and Hadoop, and use relational database to store structured feature data. The Hadoop cluster is used to store semi-/ unstructured data, make up for the shortcomings of relational databases in mass data processing and storage, and achieve high concurrent access to big data of target products and scalability of heterogeneous data.