Chaos Engineering and Situational Modeling for Fault Tolerance in High-Load Distributed Systems

Introduction. In highly loaded systems with high competitive ability and a variety of requests, the problems of high fault tolerance and competitiveness are relevant. Chaos-engineering, chaos-testing allow you to control disturbances in the system and manage its fault tolerance, increasing competitiveness.

Purpose. The research goals of the study are to analyze the tools and stages of chaos-engineering (chaos-testing) for such systems, the possibilities of social engineering in increasing fault tolerance and building a situational model of chaos-testing of such systems.

Methods. The work uses methods of analysis and synthesis of systems, mathematical and situational modeling, and optimization.

Results. Main results: 1) analytics (principles, tools, stages, metrics, social engineering) to conduct chaos-engineering, chaos-testing to manage fault tolerance of a highly competitive and highly loaded system; 2) building a situational model of fault tolerance under the conditions of the specified probabilistic distribution of "noise" and the procedure for its identification.

Conclusion. The results of the study will simplify the analysis and prediction of failures of a high-load system, conduct its load testing in conditions of chaos and high competition.