scholarly journals Distributed Load Balancing Model for Grid Computing

Author(s):  
Belabbas Yagoubi ◽  
Meriem Meddeber

International audience Most of the existing load balancing strategies were interested in distributed systems which were supposed to have homogeneous resources interconnected with homogeneous and fast networks. For Grid computing, these assumptions are not realistic because of heterogeneity, scalability and dynamicity characteristics. For these environments the load balancing problem is then a new challenge presently for which many research projects are under way. In this perspective, our contributions through this paper are two folds. First, we propose a distributed load balancing model which can represent any Grid topology into a forest structure. After that, we develop on this model, a load balancing strategy at two levels; its principal objectives : the reduction of average response time of tasks and their transferring cost. The proposed strategy is naturally distributed with a local decision, which allows the possibility of avoiding use of wide area communication network. La plupart des stratégies d’équilibrage de charge existantes se sont intéressées à des systèmes distribués supposés avoir des ressources homogènes interconnectées à l’aide de réseaux homogènes et à hauts débits. Pour les grilles de calcul, ces hypothèses ne sont pas réalistes à cause des caractéristiques d’hétérogénéité, de passage à l’échelle et de dynamicité. Pour ces environnements, le problème d’équilibrage de charge constitue donc, un nouveau défi pour lequel plusieurs recherches sont actuellement investies.Notre contribution dans cette perspective à travers ce papier est double: premièrement, nous proposons un modèle distribué d’équilibrage de charge, permettant de représenter n’importe quelle topologie de grille en une structure de forêt. Nous développons ensuite sur ce modèle, une stratégie d’équilibrage à deux niveaux ayant comme principaux objectifs la réduction du temps de réponse moyen et le coût de transfert de tâches. La stratégie proposée est de nature distribuée avec une prise de décision locale, ce qui permettra d’éviter le recours au réseau de communication à large échelle.

Author(s):  
Bellabas Yagoubi

International audience In order to get a better performance in distributed systems, load balancing problem has been extensively studied in recent years. Most of existing works focus on traditional systems where resources are generally homogeneous, like clusters. For grid infrastructures, this assumption is not totally true because resources of a grid are highly heterogeneous. Hence, load balancing problem for grid computing is a new challenge for scientists. In this paper, we propose a tree-based representation model for grid computing, over which we develop a hierarchical load balancing strategy. The main characteristics of this strategy can be summarized as follows:(i) It uses a task-level load balancing; (ii) It privileges local tasks transfer to reduce communication costs; (iii) It is a distributed strategy with local decision making. Afin d’obtenir de meilleures performances dans les systèmes répartis, le problème d’équilibrage de charge a été intensivement étudié ces dernières années. La plupart des travaux existants se sont intéressés à des systèmes qui sont plus ou moins homogènes et trouvent quelques difficultés à s’adapter aux caractéristiques des nouvelles infrastructures telles que les grilles de calcul, qui présentent un degré d’hétérogénéité assez élevé. Pour cela, il faut soit adapter, soit définir de nouvelles stratégies d’équilibrage pour ces infrastructures. Dans cette perspective, nous proposons un modèle arborescent de représentation d’une grille de calcul, sur lequel nous développons une stratégie hiérarchique d’équilibrage de charge. Les caractéristiques principales de la stratégie proposée peuvent être résumées comme suit: (i) C’est une stratégie d’équilibrage au niveau des tâches; (ii) Elle favorise un transfert local de tâches dans le but de réduire les coûts de communication; (iii) C’est une stratégie distribuée avec prise de décision locale


2013 ◽  
Vol 11 (9) ◽  
pp. 2975-2986 ◽  
Author(s):  
Said Fathy El-zoghdy

Computational grids have a huge number of diverse and scattered resources that are used in handling complex problems. A decent load balancing methodology is needed to utilize grid resource by efficiently distributing tasks, for execution, on available computing nodes.Ant colony is a major and popular method for approximate optimization. It works by simulating the actual ant‟s demeanor in detecting the best path for the resources of food. This research paper employs ant colony optimization in proposing a load balancing technique for computational grids. The performance of the suggested technique is computed, evaluated and compared with that of a Random Distributed Load Balancing technique using simulation. The achieved results reveal that the suggested technique enhances the task average response time. It reveals also that the enhancement ratio progressively rises up as the system‟s load rises up till the load come to be mild where the best enhancement ratio is achieved. Immediately after that, the enhancement ratio declines steadily as the system‟s load rises up till the system becomes saturated.


Author(s):  
ZAINAL ABIDIN ◽  
Tutuk Indriyani ◽  
Danang Haryo Sulaksono

Client’s request for traffic problems is so huge that causes the single server difficult in handling the traffic load. Therefore, the system of load balancing is required as it is a technique to equally distribute the traffic load on the two or more connection lines so that the traffic can run optimally. Thus, the load balancing is crucial to implement by using Modified Weighted Round Robin-Retrieve Packet on the Software-Defined Networking. Based on the parameter of average response-time in time limits 0.1, 0.2, and 0.3 seconds, the scores were 0.016-0.04, 0.02-0.04, and 0.014-0.032 seconds consecutively. Based on the parameter of data transaction per second in time limits 0.1; 0.2, and 0.3 seconds, the scores respectively were 49.614-111.306, 41.678-107.032, and 37.806-102.84 data transaction/second. 


2015 ◽  
Vol 5 (3) ◽  
pp. 795-800 ◽  
Author(s):  
S. F. Issawi ◽  
A. Al Halees ◽  
M. Radi

Cloud computing is a recent, emerging technology in the IT industry. It is an evolution of previous models such as grid computing. It enables a wide range of users to access a large sharing pool of resources over the internet. In such complex system, there is a tremendous need for an efficient load balancing scheme in order to satisfy peak user demands and provide high quality of services. One of the challenging problems that degrade the performance of a load balancing process is bursty workloads. Although there are a lot of researches proposing different load balancing algorithms, most of them neglect the problem of bursty workloads. Motivated by this problem, this paper proposes a new burstness-aware load balancing algorithm which can adapt to the variation in the request rate by adopting two load balancing algorithms: RR in burst and Random in non-burst state. Fuzzy logic is used in order to assign the received request to a balanced VM. The algorithm has been evaluated and compared with other algorithms using Cloud Analyst simulator.  Results show that the proposed algorithm improves the average response time and average processing time in comparison with other algorithms.


2021 ◽  
Vol 11 (4) ◽  
pp. 100-112
Author(s):  
Poonam Nandal ◽  
Deepa Bura ◽  
Meeta Singh ◽  
Sudeep Kumar

In today's world, the IT industry is emerging day by day; therefore, the need for storage and computing is increasing multifold. Cloud computing has transformed the IT sector to much greater heights by virtualizing the systems, thereby reducing cost of the hardware to greater extent. Cloud computing is based on the pay as per use policy. Due to the exponential growth in cloud computing, users demand supplementary services and improved results which makes load balancing a major challenge. Load balancing distributes the workload across multiple nodes to optimize the performance of the system. Various load balancing algorithms exist to provide better resource utilization. This paper gives a brief analysis of load balancing algorithms and also compared these algorithms on the basis of certain metrics like average response time, processing cost, and data servicing time.


Over the past few years, there has been keen research interest in load balancing and task scheduling in the cloud as the extensive amount of data that is stored in the server leads to significantly increased load. This can be resolved by using a hybrid algorithm in which the honeybee behavior algorithm’s advantages are integrated with fuzzy logic to conduct task scheduling and as well as balancing in the cloud. The design of this hybrid algorithm aims to enhance prior approaches. It is developed as per ABC and merges the important QoS factors along with power consumption so that the power that virtual machines (VMs) consume on the host can be precisely assessed, thereby ensuring efficient load balancing algorithm. The present study aims to evaluate the VMs’ power consumption by taking into account crucial QoS factors for selecting which host and virtual machine will be best suited for receiving the task. CloudSim was used to simulate the ILBA_HB algorithm. In terms of makespan, average response time, and degree of imbalance, the performance of the ILBA HB algorithm is compared to that of the LBA HB and HBB-LB algorithms. According to the results, the proposed algorithm outperformed LBA_HB and HBB-LB.


Sign in / Sign up

Export Citation Format

Share Document