Analyzes of the Distributed System Load with Multifractal Input Data Flows

TL;DR

This paper introduces a multifractal-based load balancing method for distributed systems, improving resource utilization and reducing server imbalance by accounting for complex traffic flow properties.

Contribution

It presents a novel approach that incorporates multifractal characteristics of input data flows to enhance load balancing in distributed systems.

Findings

Multifractal properties significantly affect system imbalance.

The proposed method reduces load deviation among servers.

Improves overall system performance and processing speed.

Abstract

The paper proposes a solution an actual scientific problem related to load balancing and efficient utilization of resources of the distributed system. The proposed method is based on calculation of load CPU, memory, and bandwidth by flows of different classes of service for each server and the entire distributed system and taking into account multifractal properties of input data flows. Weighting factors were introduced that allow to determine the significance of the characteristics of server relative to each other. Thus, this method allows to calculate the imbalance of the all system servers and system utilization. The simulation of the proposed method for different multifractal parameters of input flows was conducted. The simulation showed that the characteristics of multifractal traffic have a appreciable effect on the system imbalance. The usage of proposed method allows to distribute requests across the servers thus that the deviation of the load servers from the average value was minimal, that allows to get a higher metrics of system performance and faster processing flows.