The Liquid Model Load Balancing Method

Henrich, Dominik

doi:10.1080/10637199608915558

Cited by 9 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The fluid model suggested in [1,18,19,29] is adopted to approximate the system dynamics. In the fluid model, a processor is approximated by a liquid vessel in which the level of liquid represents to the processor's utilization.…”

Section: Dynamic Model Of Dlbmentioning

confidence: 99%

Distributed Utilization Control for Real-Time Clusters with Load Balancing

Wang

et al. 2006

2006 27th IEEE International Real-Time Systems Symposium (RTSS'06)

View full text Add to dashboard Cite

Recent years have seen rapid growth of online services that rely on large-scale server clusters to handle high volume of requests. Such clusters must adaptively control the CPU utilizations of many processors in order to maintain desired soft real-time performance and prevent system overload in face of unpredictable workloads. This paper presents DUC-LB, a novel distributed utilization control algorithm for cluster-based soft real-time applications. Compared to earlier works on utilization control, a distinguishing feature of DUC-LB is its capability to handle system dynamics caused by load balancing, which is a common and essential component of most clusters today.

show abstract

Section: Dynamic Model Of Dlbmentioning

confidence: 99%

Distributed Utilization Control for Real-Time Clusters with Load Balancing

Wang

et al. 2006

2006 27th IEEE International Real-Time Systems Symposium (RTSS'06)

View full text Add to dashboard Cite

show abstract

“…It is possible to reduce the expected size of the global imbalance by employing non-deterministic algorithms such as the one presented in [9], although these types of formulations do not preclude pathological cases that keep the system from balancing and also may result in a balanced system becoming unbalanced. The algorithm presented in [10], which shares some commonalities with this work, is able to achieve a globally balanced distribution but is specific to applications involving a regular grid of parallel processors and does not address the delays, asynchronism, and distributed calculation present in our model. A final difference between the usual loadbalancing framework and our problem is that in traditional load-balancing systems there is the desire that the occurrence and volume of load transfer events should diminish as load imbalances become small.…”

Section: Introductionmentioning

confidence: 99%

Decentralized redistribution for cooperative patrol

Moore

Passino

2007

Intl J Robust & Nonlinear

View full text Add to dashboard Cite

SUMMARYThis paper addresses the problem of enabling a group of autonomous vehicles to effectively patrol an environment significantly larger than their communication and sensing radii. Our formulation uses an a priori spatial decomposition of the environment into smaller areas in order to provide a framework for the allocation of vehicles to different parts of the environment. We develop a distributed cooperative control algorithm that transfers vehicles between these areas based on only local information and prove that it achieves the proper environment-wide distribution of vehicles within a finite time interval. Various applications are discussed and simulations are included to illustrate convergence dynamics as well as to quantify practical performance as a function of problem parameters.

show abstract

“…Various stochastic techniques such as randomized methods and simulated annealing and algorithms based on evolutionary algorithms have been researched extensively in the literature [18,19]. Load balancing using the models inspired from physics has also been studied in [8,9,11].…”

Section: Related Workmentioning

confidence: 99%

A physical particle and plane framework for load balancing in multiprocessors

Imani¹,

Sarbazi-Azad

2006

Proceedings 20th IEEE International Parallel &Amp; Distributed Processing Symposium

View full text Add to dashboard Cite

Different models for load balancing have been proposed before, each of which has its own features and advantages when considered for a specific scenario. Yet, nearly all of the existing techniques have assumed an oversimplified model of the system which is often not the case of the real world. In this paper, a new gradient based algorithm for dynamic load balancing on multiprocessors is proposed. This algorithm is an analogy of a classical physical model of a Particle & Plane system which operates based on the classic laws of physics dictated by the nature.

show abstract

The Liquid Model Load Balancing Method

Abstract: Abstract

Cited by 9 publications

References 13 publications

Distributed Utilization Control for Real-Time Clusters with Load Balancing

Distributed Utilization Control for Real-Time Clusters with Load Balancing

Decentralized redistribution for cooperative patrol

A physical particle and plane framework for load balancing in multiprocessors

Contact Info

Product

Resources

About