Choosing a data versioning model for designing information systems

One of the common tasks faced by developers of information systems focused on working with data is the support of versioning and history of storing records. Versioning approaches that exist today are applicable to various application scenarios: online data processing, analytical reporting, distributed storage maintenance, etc. and do not depend directly on the used database. To select the optimal versioning model, it is necessary to determine the key requirements for the system in terms of versioning operations and evaluate the effectiveness of their implementation within each of the available mechanisms. This article provides an overview of these mechanisms, their characteristic differences, advantages and disadvantages, as well as examples of use. A method is proposed for comparative evaluation of these mechanisms on a test relational database using a series of operations on versions. This technique was used by the authors in the design and development of the integrated information system. Based on the initial requirements for the system, the characteristics of the models, and the obtained evaluation results, an analysis of the effectiveness of the studied mechanisms was carried out. The result of the analysis is the choice of the most optimal model in terms of the speed of versioning operations, data integrity, and flexibility in working with attributes.