Formalized selection of open-source database management systems for MES import substitution: an architectural approach and industrial operation criteria

Abstract

This paper addresses the problem of import substitution of the data storage subsystem at the Manufacturing Execution System (MES) level of an industrial enterprise through migration from a proprietary database management system (DBMS) to an open-source solution. The relevance of such migration is substantiated by the increasing volume of production data and the growing requirements for fault tolerance, observability, and integration with modern industrial analytics tools. A formalized approach to DBMS selection is proposed based on a set of technical and operational criteria, including transaction reliability, fault tolerance, backup and recovery capabilities, scalability, time-series data processing efficiency, manageability, compatibility with enterprise IT architecture, and predictable behavior under load. Using PostgreSQL as a case study, it is demonstrated that import substitution should not be reduced to a simple replacement of a software product. Instead, it requires a revision of architectural principles, separation of development, testing, and production environments, implementation of replication mechanisms, Write-Ahead Logging (WAL) archiving, data partitioning, and monitoring tools. The practical significance of the study lies in the applicability of the proposed criteria and evaluation approach to the modernization of existing MES solutions and the design of new industrial information platforms based on open technologies.