OPTIMIZING PARAMETERS OF UNIVERSITY CLOUD SYSTEMS TO ENSURE RELIABILITY AND SECURITY IN IDENTITY DECENTRALIZATION

Abstract

It is shown that microservice architecture (MSA) of cloud services (CS) of a private university cloud is an approach to software development (SW), according to which an application is built from small ones independent services that interact with each other through an APIs, with each service responsible for execution specific functionality and can be deployed and scaled separately, thereby enabling the creation of flexible and scalable applications that can quickly adapt to changing conditions and workloads for each task of using microservice architecture. A review and analysis of previous studies devoted to the use of MCA in cloud computing (CC) and CS was conducted, which showed that CMA is widely used in CC and CS due to its flexibility, scalability and high availability. It is shown that the decentralization of identity will make it possible to increase the security and privacy of user data, because each service in the MSA CS can have its own access and identity management system, which, on the one hand, leads to a number of advantages, such as scalability, flexibility and resistance to failures, but on the other hand, the decentralized nature of the MSA also creates problems with identity management. The model for simulating the system load factor during the MSA was further developed. The solution proposed in the work, unlike existing solutions, takes into account the limitations of decentralized identity management (i.e., DID). A software solution using the Python algorithmic language for the implementation of a mathematical model for the simulation of CS parameters is described. The model, among other things, takes into account the weights (importance) of applications entering the decentralized identity management system.