DEVELOPMENT OF A BASIC MODEL FOR CHOOSING AN INFORMATION SYSTEM CLOUD MIGRATION STRATEGY

Abstract

Cloud migration of information systems is currently one of the critical transitional stages of enterprise development on the way to digital transformation. The problem of choosing a strategy for cloud migration of the information system at the pre-migration stage is considered as a task of achieving the target characteristics of quality and productivity, an acceptable level of costs and manageability of implementation in conditions of limited initial data. It is determined that the existing approaches to migration planning are often diverse and fragmented. The formalization of the choice of strategy as a comparison of alternatives in the common space of key performance indicators is proposed. The target indicator profile is introduced as a set of minimum acceptable thresholds after migration, as well as important weights that reflect the priorities of stakeholders and allow you to explicitly capture which requirements are critical. The suitability of each alternative as a measure of failure to achieve the target thresholds for indicators where the predicted state did not meet the requirements is determined. The compromise between the achievement of the target profile and the complexity of implementation is considered, which made it possible to control the balance between the expected improvements and resource costs for the transformation of the information system. Experimental verification of the model on a given set of input data was performed, which confirmed the determinism and reproducibility of the results. With fixed initial values of indicators, normalization rules, threshold goals, weights of importance and compromise parameters, the model formed an unambiguous ranking of strategies and made it possible to identify which indicators formed the main deficit and in which directions changes were needed to reduce it. It is substantiated that such a formulation strengthened the transparency and reasoning of the choice of a cloud migration strategy. Further areas of research are defined as calibration of model parameters on empirical examples, analysis of the sensitivity of results to weights and thresholds, considering the uncertainty of indicator forecasts, as well as expanding the mechanisms of scenario comparison to increase the portability of recommendations for different domains and types of information systems.