Divisible Load Research Articles

EDLT: An Extended DLT to Enhance Load Balancing in Cloud Computing

In recent years, cloud computing has gained prominence and most companies are applying some kind of cloud computing technology in their company or part of it. Growing number of clients who are willing to use cloud services has made load balancing an eminent challenge in this field. A general approach to load balancing is the application of Divisible Load Theory (DLT). In DLT, the workload is divided among master systems. For their turn, master systems divide the load among their slave systems. This article presents an enhanced technique for load balancing based on DLT. Simulated results indicate that the expanded DLT reduces measurement/report time and shows an improved performance at a lower failure rate.

Communication scheduling in data gathering networks with limited memory

In this paper scheduling communications in data gathering networks is analyzed. We study collecting information by a set of sensors, each of which stores the data in its memory buffer and then passes them to a base station. The network lifetime ends as soon as the first node is out of memory. We use a divisible load model to propose a communication scheduling algorithm that maximizes the system lifetime, and hence, the total amount of gathered information. The influence of the sensor communication rates on the obtained schedules is exposed in a series of computational experiments.

Divisible Load Theory Based Active-Sleep Workload Assignment Schemes for Wireless Sensor Networks

Wireless sensor networks (WSNs) have been widely applied in many monitoring and surveillance processes. Due to the limited onboard energy resources, the operation of a WSN is severely hindered by its shortened lifespan arising from energy depletion. In order to alleviate this problem, and with the possibility of partitioning sensing workloads, the divisible load theory (DLT) can be adopted to derive a proper workload assignment scheme. However, special considerations have to be paid for its generic applicability in feasible workload assignment for WSNs. In this paper, an examination of DLT based WSN operations including the effect of assignment, measurement, and report times is presented, and the problem of negative workloads inherently generated in some schemes is revealed. Furthermore, by making use of the negativity phenomenon, an active-sleep scheme is proposed for the WSN such that sensor energy consumptions can be reduced and consequently extend the WSN operation lifespan. Specifically, sensors with smaller amount of residual energies are put into the sleep mode when the assigned workloads are negative. On the other hand, positive workloads are normalized and reassigned to sensors with larger amount of onboard energies. Simulation studies are carried out to demonstrate the negative workload phenomenon, and satisfactory performances of the proposed active-sleep scheme are verified.

Implementation and performance evaluation of a scheduling algorithm for divisible load parallel applications in a cloud computing environment

SummaryCloud computing is an emerging technology in which information technology resources are virtualized to users in a set of computing resources on a pay‐per‐use basis. It is seen as an effective infrastructure for high performance applications. Divisible load applications occur in many scientific and engineering applications. However, dividing an application and deploying it in a cloud computing environment face challenges to obtain an optimal performance due to the overheads introduced by the cloud virtualization and the supporting cloud middleware. Therefore, we provide results of series of extensive experiments in scheduling divisible load application in a Cloud environment to decrease the overall application execution time considering the cloud networking and computing capacities presented to the application's user. We experiment with real applications within the Amazon cloud computing environment. Our extensive experiments analyze the reasons of the discrepancies between a theoretical model and the reality and propose adequate solutions. These discrepancies are due to three factors: the network behavior, the application behavior and the cloud computing virtualization. Our results show that applying the algorithm result in a maximum ratio of 1.41 of the measured normalized makespan versus the ideal makespan for application in which the communication to computation ratio is big. They show that the algorithm is effective for those applications in a heterogeneous setting reaching a ratio of 1.28 for large data sets. For application following the ensemble clustering model in which the computation to communication ratio is big and variable, we obtained a maximum ratio of 4.7 for large data set and a ratio of 2.11 for small data set. Applying the algorithm also results in an important speedup. These results are revealing for the type of applications we consider under experiments. The experiments also reveal the impact of the choice of the platforms provided by Amazon on the performance of the applications under study. Considering the emergence of cloud computing for high performance applications, the results in this paper can be widely adopted by cloud computing developers. Copyright © 2014 John Wiley & Sons, Ltd.

Optimization Algorithm for Divisible Load Scheduling on Heterogeneous Star Networks

Scheduling divisible loads on heterogeneous distributed computing systems is addressed in this paper. The platform considered here is more general and realistic, where processors are connected in star topology with arbitrary communication and computation speeds and non- zero start-up overheads. A new optimization algorithm, called WX-GA, is proposed to tackle the following four issues: (1) How many processors are needed in computation? (2) Finding the optimal distribution sequence among processors. (3) How much the load fraction should be assigned on each processor? (4) When workload is large enough, what is the sufficient and necessary condition for the minimum processing time? Finally, the experimental results indicate the efficiency and effectiveness of the proposed algorithm.

Multi-source task scheduling in grid computing environment using linear programming

In grid computing environment the workload can be submitted by various grid users. Allocating a load to the resources from various grid users is a challenging task in grid scheduling process. So we need a better resource allocation framework to schedule the tasks from various sources to the dynamic resources. This paper introduces a new divisible load scheduling framework to map the task to the distributed resources in grid computing environment. We have developed a mathematical model to allocate tasks to resource when the task is submitted by various grid users sources with the aim of minimising the completion time of scheduling process with satisfying the budget allotted by the grid user based on divisible load theory. This model helps to estimate the resource usage cost of each source under the fluctuation of resource availability and the performance evaluation of three resource homogeneous resources and the link capacity is presented.

Resource information sharing in multi-domain optical grids

Providing interoperability in multi-domain optical grids requires sharing resource information among different domains. Sharing full resource information may not be appropriate due to security and business confidentiality reasons. We propose resource information sharing approach based on information aggregation and abstraction for both computing and networking resources. Information aggregation and abstraction affects the overall accuracy of information representation. This calls for more accurate information abstraction techniques. We propose three abstraction techniques: direct formulas, historical data and real-time measurements. The performance of the proposed information sharing approach and the abstraction techniques is evaluated by studying their impact on scheduling divisible load grid applications. Simulation results highlight the advantages of the proposed approach specially for data intensive applications. By using the proposed abstraction techniques, the same performance could be achieved as using large volumes of exact information.

Comprehensive Review on Divisible Load Theory: Concepts, Strategies, and Approaches

There is extensive literature concerning the divisible load theory. The divisible load theory is mainly applied for scheduling in the area of distributed computing. It is based on the fact that the load can be divided into some arbitrarily independent parts, in which each part can be processed independently by a processor. This paper reviews the literature concerning the divisible load theory, while focusing on the details of the basic concepts, approaches, strategies, typologies, and open problems.

NEW MODEL AND GENETIC ALGORITHM FOR DIVISIBLE LOAD SCHEDULING IN HETEROGENEOUS DISTRIBUTED SYSTEMS

The problem of divisible load scheduling in network based heterogeneous distributed systems is addressed in this paper, where a general platform is considered, and the communication is in non-blocking message receiving mode, moreover, the communication speeds, computation speeds, start-up overheads and workload size are arbitrary. To solve the problem efficiently, we set up an optimization model which can effectively tackle the following three issues: (1) how many and which processors are required in computation; (2) in which order the load fractions are distributed to processors; (3) how much the load fraction should be distributed to each processor. For this model, a novel genetic algorithm is proposed, and the convergence of the proposed algorithm to a globally optimal solution with probability one is proved. Finally, the experiments on several examples indicate the efficiency and effectiveness of the proposed algorithm.

Energy trade-offs analysis using equal-energy maps

Energy consumption has become a factor limiting further progress of supercomputing. Grasping the relationships determining the transformation of energy into computations is often difficult. Therefore, we propose a new method of visualizing the relationships as two-dimensional maps similar to isotherms or isobars in weather maps. Complex models of energy consumption are projected onto two-dimensional maps with isolines representing points of equal energy consumption. As an illustration of the concept, we present isoenergy maps for three models of parallel computations. The first two models are derived from Amdahl’s and Gustafson’s laws of parallel performance. The third model applies to divisible loads representing data-parallel computations in distributed systems. Equal-energy maps for the three models provide qualitative and quantitative insights into the interactions between certain energy-saving approaches, and their consequent limitations.

Su1219 Accurate Determination of Serum Adalimumab and Anti-Adalimumab Antibodies Levels During Maintenance Therapy for Crohn's Disease

Min, Veeravalli, and Barlas have proposed strategies to minimize the overall execution time of one or several divisible loads on a heterogeneous linear network, using one or more installments [Han Min Wong, Bharadwaj Veeravalli, Scheduling divisible loads on heterogeneous linear daisy chain networks with arbitrary processor release times, IEEE Trans. Parallel Distrib. Syst. 15 (3) (2004) 273–288; Han Min Wong, Bharadwaj Veeravalli, Gerassimos Barlas, Design and performance evaluation of load distribution strategies for multiple divisible loads on heterogeneous linear daisy chain networks, J. Parallel Distrib. Comput. 65 (12) (2005) 1558–1577]. We show using a very simple example that their approach does not always produce a solution and that, when it does, the solution is often suboptimal. We also show how to find an optimal scheduling for any instance, once the number of installments per load is given. Finally, we formally prove that under a linear cost model, as in both the above-mentioned references, an optimal schedule has an infinite number of installments. Therefore such a cost model should not be used to design practical multi-installment algorithms.

Requirement-aware strategies for scheduling real-time divisible loads on clusters

This paper investigates the real-time scheduling problem for handling heterogeneous divisible loads on cluster systems. Divisible load applications occur in many fields of science and engineering. Such applications can be easily parallelized in a master–worker fashion, but pose several scheduling challenges. We consider divisible loads associated with deadlines to enhance quality-of-service (QoS) and provide performance guarantees in distributed computing environments. In addition, since the divisible loads to be performed may widely vary in terms of their required hardware and software, we capture the loads’ various processing requirements in our load distribution strategies, a unique feature that is applicable for running proprietary applications only on certain eligible processing nodes. Thus in our problem formulation each load can only be processed by certain processors as both the loads and processors are heterogeneous. We propose scheduling algorithms referred to as Requirements-Aware Real-Time Scheduling (RARTS) algorithms, which consist of a novel scheduling policy, referred to as Minimum Slack Capacity First (MSCF), and two multi-round load distribution strategies, referred to as All Eligible Processors (AEP) and Least Capability First (LCF). We perform rigorous performance evaluation studies to quantify the performance of our strategies on a variety of scenarios.

Adaptive Indexed Divisible Load Theory for Wireless Sensor Network Workload Allocation

Energy depletion in wireless sensors is a major obstacle for a wireless sensor network (WSN) to operate over an extended period of time. This problem can be extenuated by minimizing the need for high-power transmission from sensors to the master processor. Sensors could be arranged in clusters, and their sensing workloads are properly determined for minimal energy consumption during the sensing and result reporting stages. The divisible load theory (DLT) is applied here to obtain optimal allocation of sensor workloads taking into account the balance of energy used such that the failure of the first sensor can be delayed. Since standard DLT assumes an ordered indexing of the sensors, its direct application in WSNs may result in unbalanced energy usage. Adaptive indexing schemes with the application of DLT, adaptive indexed divisible load theory (AIDLT), are thus proposed to redefine the indices of sensors in each sensing round while calculating the assigned workload portions. Furthermore, adaptations based on transmission distances, sensor residual energies, double ranking of distances with residual energies, and randomized sensor identifications are formulated and evaluated. Simulation results on a cluster of sensors have shown that adaptation based on residual energies outperforms the other indexing schemes while the randomization scheme is the simplest.

Cost-based Multi-QoS Job Scheduling Using Divisible Load Theory in Cloud Computing

The advent of cloud computing as a new model of service provisioning in distributed systems, encourages researchers to investigate its benefits and drawbacks in executing scientific applications such as workflows. In this research, we attempt to investigate the use of a Divisible Load Theory (DLT) to design efficient strategies to minimize the overall processing time for scheduling jobs in compute cloud environments. We consider homogeneous processors in our analysis and we derive a closed-form solution for the load fractions to be assigned to each processors. Our analysis also attempts to schedule the jobs such a way that cloud provider can gain maximum benefit for his service and Quality of Service (QoS) requirement user's job. Finally, we quantify the performance of the strategies via rigorous simulation studies.

Bandwidth‐aware divisible task scheduling for cloud computing

SUMMARYTask scheduling is a fundamental issue in achieving high efficiency in cloud computing. However, it is a big challenge for efficient scheduling algorithm design and implementation (as general scheduling problem is NP‐complete). Most existing task‐scheduling methods of cloud computing only consider task resource requirements for CPU and memory, without considering bandwidth requirements. In order to obtain better performance, in this paper, we propose a bandwidth‐aware algorithm for divisible task scheduling in cloud‐computing environments. A nonlinear programming model for the divisible task‐scheduling problem under the bounded multi‐port model is presented. By solving this model, the optimized allocation scheme that determines proper number of tasks assigned to each virtual resource node is obtained. On the basis of the optimized allocation scheme, a heuristic algorithm for divisible load scheduling, called bandwidth‐aware task‐scheduling (BATS) algorithm, is proposed. The performance of algorithm is evaluated using CloudSim toolkit. Experimental result shows that, compared with the fair‐based task‐scheduling algorithm, the bandwidth‐only task‐scheduling algorithm, and the computation‐only task‐scheduling algorithm, the proposed algorithm (BATS) has better performance. Copyright © 2012 John Wiley & Sons, Ltd.

Greedy Signature Processing with Arbitrary Location Distributions: A Divisible Load Framework

The optimal partition of a huge, linear (flat) file among processing nodes in a network to minimize the time to search for signatures of interest in the file is considered. First, an expression is developed for the expected time of finding the kth signature (including the last signature) of K signatures for a uniform distribution of signatures in the file. Secondly, for a single signature we propose processing data in the file in order from that with the most probability mass (i.e., data with the most a priori likelihood of containing the signature) to that with the least probability mass in a "greedy" manner to speed processing time. Applications of this work include radar, sensors, image processing, and search.

Efficient real-time divisible load scheduling

Providing QoS and performance guarantees to arbitrarily divisible loads has become a significant problem for many cluster-based research computing facilities. While progress is being made in scheduling arbitrarily divisible loads, current approaches are not efficient and do not scale well. In this paper, we propose a linear algorithm for real-time divisible load scheduling. Unlike existing approaches, the new algorithm relaxes the tight coupling between the task admission controller and the task dispatcher. By eliminating the need to generate exact schedules in the admission controller, the algorithm avoids high overheads. We also proposed a hybrid algorithm that combines the best of our efficient algorithm and a previously best-known approach. We experimentally evaluate the new algorithm. Simulation results demonstrate that the algorithm scales well, can schedule large numbers of tasks efficiently, and performs similarly to existing approaches in terms of providing real-time guarantees.

Novel Adaptive Job Scheduling Algorithm on Heterogeneous Grid Resources

Grid provides an infrastructure for sharing geographically distributed heterogeneous resources to process many applications and mainly used for solving scientific problems that requires more computation time. Problem statement: Grid is a dynamic environment, where the resources may join or leave the environment at any time and the jobs also arrives at different intervals of time. To meet the demands and requirements of the dynamic environment, to maximize the resource utilization and to minimize the makespan an effective grid scheduling technique is needed. Approach: We propose grid architecture as a collection of clusters with multiple worker nodes in each cluster. We propose a new scheduling algorithm Novel Adaptive Decentralized Job Scheduling Algorithm (NADJSA) that applies both Divisible Load Theory (DLT) and Least Cost Method (LCM) and also considers the user demands. Results: The proposed Novel Adaptive Decentralized Job Scheduling Algorithm is compared with the Decentralized Hybrid Job Scheduling Algorithm. Conclusion: The proposed Novel Adaptive Decentralized Job Scheduling Algorithm minimizes the makespan, improves the resource utilization and satisfies the user demands and well suits for the grid environment.

Techniques for Load Balancing in Wireless LAN

This suggests new strategies for balancing load in a wireless network connected in star topology. The loads are assigned to each processor using divisible load theory & Different techniques [II], [III], [IV], and [V]. Divisible load theory suggests that a load can be divided arbitrarily such that each fraction of the load can be independently assigned and computed in any processor present in the network. Wireless networks are connected in such a manner that they as assemble a distributed system most of the times, which makes load balancing an important technique to maximize the throughput from the system. A wireless sensor network generally consists of a base station (or Gateway) which communicates with other nodes present in the network. The other nodes are used for Measuring and collecting various environmental and Intelligence related data. The network that we have considered is connected with the central node being the base station and the other nodes are used for calculation of load distributed by the central node. Load balancing involves distribution of all computational and communicational activities over two or more processors, links or any other computational devices present in the network. The main thing behind this is load balancing is to reduce the execution time of the load and to make sure that all the resources present in the system are utilized optimally. The IEEE 802.11 standard does not provide any mechanism to resolve load imbalance. To reduce this deficiency, various load balancing schemes have been designed. These techniques commonly take the approach of directly controlling the user-AP association by deploying Proprietary client software or hardware. Load balancing Features in their device drivers, AP firm wares, and WLAN cards. In these solutions, APs broadcast their load levels to users via modified beacon messages and each user chooses the least-loaded AP.

Load Balancing in Optical Grids

This collaboration of geographically distributed domains in multi-domain optical grid environment should be done in a way that maintains the privacy of each participant domain. This calls for a new load balancing hierarchical approach to deal with such environments. In this paper, we propose two load balancing techniques within this hierarchical approach: Network Aware Divisible Load Algorithm (NADLA) and Genetic Algorithm based Load Distribution (GA-LD). Both algorithms are considered as Network Aware. Simulation results show the scalability and feasibility of the proposed approach and the advantages of the two proposed techniques compared to the classical Divisible Load Algorithm (DLA). In the last years, the growing of the scientific and enterprise applications pushes grid systems to more scalable and collaborative direction. Single domain resources are becoming insufficient to deal with application requirements. Grid applications require massive amounts of data to be transferred and processed in geographically distributed sites across the grid. The grid is an inter- connected multi-domain environment where each domain consists of computational, storage and communication resources grouped together for business or administrative reasons. Each domain is independently administrated and is free to deploy different technologies. Achieving better resource utilization calls for running load balancing techniques across different grid domains. This calls for novel load balancing techniques running in multi-domain scalable architecture to achieve better resource utilization without sacrificing domain security or privacy. The proposed multi-domain optical grid architecture assumes a dynamic on-demand reconfigurable optical network instead of pre-planned statically provisioned optical network. Dynamic reconfigurable optical grids presented in different projects such as Phosphorus (1) and G-lambda Projects (2). These projects present the structure and different components of optical network deployment in grids, and how these components interact to provide the grid applications with the services and capabilities making it able to manage its grid and optical network resource. In this paper, we present load balancing techniques in multi-domain hierarchical optical grid architecture that maintains the privacy, integration and scalability requirements. The privacy of a grid domain must be maintained in for confidentiality and commercial competition. For example, internal topology information of a domain should not be revealed to other domains (3). The domain privacy could be maintained by keeping the full resource information internally, while