Methods and models of the resource allocation service in clusters with load balancing of data centers

Modern data processing centers are complex solutions for the management of enterprises and corporations, for the organization of data processing and storage systems, for the efficient distribution of software applications between available resources. The object of the research is load-balancing clusters of data processing centers containing a certain set of application servers, file servers, data storage systems, an input-output system interconnected by a switching system and communication channels. The aim of the work is to increase the efficiency of the functioning of virtualized data centers by developing methodological approaches, rational methods and models for solving the problems of load distribution by its hardware and software. The problems of constructing interrelated mathematical methods and models necessary for building specialized software that provide a solution to the problems of creating rational plans for the allocation of data center resources at a given time interval are considered. The proposed methods and models are based on the system of correct mappings of the set of query parameters for the known characteristics of the physical resources of the data center. As an optimization criterion, the parameters of maximizing system performance for a certain time interval are used.