Abstract:The dynamic load balancing algorithm based on the monitoring server load, self-similar characteristics of passing traffic have to provide a statistically uniform load distribution on servers, high performance, fault tolerance and capacity, low response time, the amount of overhead and losses was propose in work. Integrated measurement for the total imbalance level of the system were entered.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to achieve dynamic load balancing in distributed systems to cope with weakly predictable load fluctuations in modern communication networks. Specifically, the paper proposes a dynamic load - balancing algorithm based on server load monitoring and self - similar traffic characteristics, aiming to provide a statistically uniform load distribution, improve performance, fault - tolerance ability and capacity, while reducing response time, service traffic and loss.
The paper points out that traditional static planning strategies are no longer applicable in network management, especially in modern traffic engineering systems, where both online and offline methods are required to manage network loads. However, the disadvantage of offline control systems lies in their inability to respond to changes in the current load matrix, while online methods, although able to monitor network states in real - time, lack global load prediction, resulting in routing instability. Therefore, the paper proposes an adaptive mechanism that combines these two methods to solve these problems by monitoring the current situation and using prediction to reduce the frequency of re - routing.
In addition, the paper also introduces a comprehensive measurement index for evaluating the overall imbalance level of the system and the average imbalance level of each server, which helps to perform load distribution and resource scheduling more effectively.