Distributed Database Management System Service Model with Cluster Load Rebalance

The work shows the problem of horizontal scaling of database management systems when maintaining their query performance. Problems of both structured-data and semi-structured-data architecture styles are highlighted. We have shown the lack of available solutions that can scale all the software system components, distribute data for uniform and efficient usage of hardware resources, maintain both structured and non-structured data simultaneously, and provide data integrity functionality, stored procedures, and triggers. As a solution to the mentioned problem, we proposed a heterogeneous distributed database management system service model able to redistribute data in a cluster based on performance metrics. We propose a new method for the creation of new distributed objects. The method optimizes performance metrics by minimizing the objective function under specified constraints. The software implementation and the technological stack, data types, database operations, and verification process for the proposed model are described in detail.