Architectures for Processing Large Data Arrays in Distributed Systems Based on Open Source Frameworks

Currently, relational and object-relational databases have become the de facto standard for organizing data storage in most information systems. But such structures are not suitable for volumes of information that cannot be processed by the capacity of a single server. To process such big data, databases with a key-value structure are used. The article discusses the main problems that arise when creating distributed big data repositories and modern technologies that allow these problems to be leveled. There is no single universal solution related to the storage and processing of big data. When choosing technologies and implementing storage, it is important to consider together the tasks that require solutions, technical capabilities, types and regularity of necessary treatments, data features, for example, the frequency of their updates.