Comparing load balancing algorithms for distributed queueing networks

Applications of queueing network models have multiplied in the last generation, including scheduling of large manufacturing systems, control of patient flow in health systems, load balancing in cloud computing, and matching in ride sharing. These problems are too large and complex for exact solution, but their scale allows approximation. This book is the first comprehensive treatment of fluid scaling, diffusion scaling, and many-server scaling in a single text presented at a level suitable for graduate students. Fluid scaling is used to verify stability, in particular treating max weight policies, and to study optimal control of transient queueing networks. Diffusion scaling is used to control systems in balanced heavy traffic, by solving for optimal scheduling, admission control, and routing in Brownian networks. Many-server scaling is studied in the quality and efficiency driven Halfin–Whitt regime and applied to load balancing in the supermarket model and to bipartite matching in ride-sharing applications.

Download Full-text

Power-of-d-Choices with Memory: Fluid Limit and Optimality

Mathematics of Operations Research ◽

10.1287/moor.2019.1014 ◽

2020 ◽

Vol 45 (3) ◽

pp. 862-888 ◽

Cited By ~ 1

Author(s):

Jonatha Anselmi ◽

Francois Dufour

Keyword(s):

Load Balancing ◽

Fundamental Problem ◽

Asymptotic Properties ◽

Queueing Systems ◽

Queue Lengths ◽

Randomized Load Balancing ◽

Load Balancing Algorithm ◽

Load Balancer ◽

Distributed Queueing ◽

With Memory

In multiserver distributed queueing systems, the access of stochastically arriving jobs to resources is often regulated by a dispatcher, also known as a load balancer. A fundamental problem consists in designing a load-balancing algorithm that minimizes the delays experienced by jobs. During the last two decades, the power-of-d-choice algorithm, based on the idea of dispatching each job to the least loaded server out of d servers randomly sampled at the arrival of the job itself, has emerged as a breakthrough in the foundations of this area because of its versatility and appealing asymptotic properties. In this paper, we consider the power-of-d-choice algorithm with the addition of a local memory that keeps track of the latest observations collected over time on the sampled servers. Then, each job is sent to a server with the lowest observation. We show that this algorithm is asymptotically optimal in the sense that the load balancer can always assign each job to an idle server in the large-system limit. This holds true if and only if the system load λ is less than [Formula: see text]. If this condition is not satisfied, we show that queue lengths are bounded by [Formula: see text]. This is in contrast with the classic version of the power-of-d-choice algorithm, in which, at the fluid scale, a strictly positive proportion of servers containing [Formula: see text] jobs exists for all [Formula: see text] in equilibrium. Our results quantify and highlight the importance of using memory as a means to enhance performance in randomized load balancing.

Download Full-text

On the conceptual design of multi-agent system for load balancing using multi-class queueing networks

2015 Workshop on Engineering Applications - International Congress on Engineering (WEA) ◽

10.1109/wea.2015.7370128 ◽

2015 ◽

Cited By ~ 2

Author(s):

Eduyn Ramiro Lopez Santana ◽

German Andres Mendez Giraldo ◽

Gustavo Florez Becerra

Keyword(s):

Load Balancing ◽

Conceptual Design ◽

Queueing Networks ◽

Multi Agent System ◽

Agent System ◽

Multi Agent

Download Full-text

Task Scheduling Algorithm based on Resources Segregation in Cloud Environment

International Journal of Scientific Research in Science Engineering and Technology ◽

10.32628/ijsrset21841125 ◽

2018 ◽

pp. 96-101

Author(s):

Shailendra Raghuvanshi ◽

Priyanka Dubey

Keyword(s):

Load Balancing ◽

Task Scheduling ◽

Waiting Time ◽

Virtual Machines ◽

Scheduling Algorithm ◽

Cloud Environment ◽

Task Scheduling Algorithm ◽

Machine Utilization ◽

Independent Tasks ◽

Optimal Machine

Load balancing of non-preemptive independent tasks on virtual machines (VMs) is an important aspect of task scheduling in clouds. Whenever certain VMs are overloaded and remaining VMs are under loaded with tasks for processing, the load has to be balanced to achieve optimal machine utilization. In this paper, we propose an algorithm named honey bee behavior inspired load balancing, which aims to achieve well balanced load across virtual machines for maximizing the throughput. The proposed algorithm also balances the priorities of tasks on the machines in such a way that the amount of waiting time of the tasks in the queue is minimal. We have compared the proposed algorithm with existing load balancing and scheduling algorithms. The experimental results show that the algorithm is effective when compared with existing algorithms. Our approach illustrates that there is a significant improvement in average execution time and reduction in waiting time of tasks on queue using workflowsim simulator in JAVA.

Download Full-text