The paper examines the existing solutions design of various data warehouses. The main trends in the development of technologies are identified. An analysis of existing big data classifications allowed us to offer our own measure to determine the category of data. On its basis, a new classification of big data has been proposed (taking into account the CAP theorem). A description of the characteristics of the data for each class is given. The developed big data classification is aimed at solving the problems of selecting tools for the development of an ecosystem. The practical significance of the results obtained is shown by the example of determining the type of big data of actual information systems.