Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back Towards Managing of Reliability Level of Cluster Information Systems Based on Fuzzy Inference V.A. GONCHARENKO, A.D. KHOMONENKO, S.G. ERMAKOV, R. ABU KHASAN An approach to managing the reliability of cluster information systems using fuzzy inference is proposed. The stages of fuzzy inference using the Mamdani algorithm are considered and membership functions for various levels of reliability are set. The fuzzy rules for choosing methods to achieve reliability, including redundancy, control of calculations and N-version programming, are substantiated. The schemes of rules for critical and non-critical systems are presented, as well as the main methods for ensuring their reliability. N-version programming uses the majority principle of choosing the right result. The current research directions are highlighted, taking into account the interval uncertainty of the parameters of time distributions, in order to increase the accuracy of evaluating the characteristics of systems. Keywords: fuzzy inference, cluster information systems, reliability level, redundancy, computing control, N-version programming. DOI: 10.25791/pribor.2.2025.1560 Pp. 52-58.