Современные информационные технологии и IT-образование (Dec 2022)
Method of Increasing the Autonomy of Complex Distributed Computing Systems
Abstract
This paper considers the problem of increasing the autonomy of the functioning of cluster distributed computing systems (CDCS). The goal of this research work is to increase the autonomy of functioning of cluster distributed computing systems. In accordance with the goal, the article solves the following scientific task: taking into account the analysis of the application of existing cluster distributed computing systems, their architectures and features, as well as the analysis of development trends of domestic developments in this area, develop a methodology for determining and setting the main operational requirements for promising cluster distributed computing systems that increase their autonomy. Initially, the paper describes the current state of the issue of autonomy of systems, including consideration of the existing criteria for the compliance of a computer system with the concept of autonomy, as well as the classification of systems according to the degree of their autonomy, and determines the necessary list of tasks for autonomous functioning. On the basis of this, a formalized apparatus is constructed, 2 groups of tasks are identified that are necessary to solve the issue of autonomous functioning of the CDCS, a definition of an autonomous computer system is given through the probability of operability and functionality of technical, software and information support, as well as the definition of an elementary and complex object. A feature function affecting the performance of the target task of a complex object and characterizing the intensity of malfunctions and functional failures in a unit of time is presented. The function of the coefficient of autonomy of an elementary object is introduced, on the basis of which the function of the coefficient of autonomy of a complex object is determined. On the basis of the introduced definitions and functions the formalized solution of the task of increasing autonomy is realized by maximizing the complex indicator of autonomy, based on predicting the period of autonomous task execution without external interference in the process of its functioning.
Keywords