QoS-based Scheduling for Task Management in Grid Computing

Size: px
Start display at page:

Download "QoS-based Scheduling for Task Management in Grid Computing"

Transcription

1 QoS-based Scheduling for Task Management in Grid Computing Xiaohong Huang 1, Maode Ma 2, Yan Ma 1 Abstract--Due to the heterogeneity, complexity, and autonomy of wide spread Grid resources, the dynamic scheduling with QoS concerns in Grid computing becomes a challenge. This paper addresses the dynamic scheduling problem of parallel jobs with QoS demands in Grid environment. We outline a framework for the solution termed as QoS-based Scheduling Framework (QSF). QSF works based on a general adaptive scheduling heuristics that provides a fast and efficient approach to achieve load balancing inside the system with the concerns of QoS service. It is adaptive to the changes of the number of available resources and the quality of the resources. Meanwhile, one novel scheduling algorithm, namely Minimum Scheduling Hole First (MSHF), has been proposed to give suitable match between the eligible jobs and the available hosts. Extensive experimental studies have been conducted to verify the effectiveness of the scheduling mechanisms and the performance. The experimental results show that by integrating the proposed the scheduling algorithm MSHF into the scheduler QSF, it can produce a significant performance gain. Keywords--Grid Computing, Scheduling, Quality of Service (QoS) T I. INTRODUCTION HE current computational power demands and constraints of organizations have led to a new type of collaborative computing environment called Grid computing [1]. Grid computing is a new type of parallel and distributed computing. Currently, considerable research efforts in Grid computing have focused on the issues of security, resource scheduling and complex execution frameworks, etc. However, there are still many unsolved problems, in particular, those related to the management of the processing load inside the Grid system. Job scheduler is an important component which is responsible for the load balancing in a distributed computing system. So far, many scheduling schemes have been proposed [2-5]. However, most of them only address the problems that how to develop a schedule of processing jobs on a set of heterogeneous hosts that minimize the time required to execute the given jobs. Only recently, there are considerations given to a fundamental problem prevailing in service-oriented This work was supported by China Next Generation Internet (CNGI) project Large scale high performance grid application IPv6 based (CNGI A) and Co-struction project of Beijing Committee of Education (SYS ). Xiaohong Huang Author is with School of Computer Science and Technology, Beijing University of Posts and Telecommunications, Beijing, China ( huangxh@buptnet.edu.cn) Maode Ma Author is with School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore ( Maode_Ma@pmail.ntu.edu.sg) Yan Ma Author is with School of Computer Science and Technology, Beijing University of Posts and Telecommunications, Beijing, China ( mayan@bupt.edu.cn) architecture to provide QoS services. Providing QoS service is one of the primary goals of the Grid approach. Without QoS concerns, a considerable degradation in performance and lower efficiency of the Grid computing could be incurred when a large number of jobs are issued for sharing Grid resources. One of the important aspects of jobs with high QoS requirements is that the execution of a job must be finished within a given amount of time. This time period is referred to as a job deadline. In a real-time Grid system, the principle task for the scheduler is to maximize the number of jobs that are completed before their respective deadlines. This paper proposes a semi-static, adaptive framework, termed QoS-based Scheduling Framework (QSF). QSF has its unique features following. 1) QSF is a QoS-based scheduler, which considers application and their QoS constraints to get a suitable match between applications and resources; 2) Communication overhead has been considered as an important parameter in the scheduling, which makes QSF is a good candidate for QoS services in Grid environment, especially, when the communication overhead is significant. This paper also provides an insight into scheduling algorithms along with the schedulers. In order to map the jobs to hosts, several static algorithms, namely trivial and best-fit algorithm, have been proposed in [4] and [5]. However, those solutions neglect the factor of communication overhead, which may cause large scheduling gap. The scheduling gap or hole, in this paper, has been defined as the difference between the moment when a job starts execution on a host and the moment when the host becomes idle. In order to minimize scheduling holes, one novel scheduling algorithm, namely Minimum Scheduling Hole First (MSHF), has been proposed to work the proposed scheduler, QSF. The organization of this paper is as follows. In Section 2, system model under the study will be presented. In Section 3, the proposed scheduler as well as the scheduling algorithm and their advantages are illustrated. In Section 4, experimental results are shown and discussed to present the performance of the solution. Finally, Section 5 concludes the paper. II. SYSTEM MODEL The Grid computing system considered in this study is composed of two sets of hosts. Each set of hosts has different QoS characteristics. Moreover, the Grid environment is heterogeneous and each host has different processing power. The Grid can be considered as the one with one agent, i.e., scheduler, with other hosts as work stations. The agent in each set is to manage the execution of parallel jobs within the group. And the agent will make decisions to distribute the jobs and assign the jobs to the hosts proportionally to the power of each host. 406

2 The application model, which describes the relationship among the jobs exhibits precedence constraints among the parallel jobs. Some jobs cannot be executed until other jobs have been completed. Each job is ineligible until all of its precedence constraints have been satisfied. The application and Grid models in this paper can be implemented by most of the famous Grid middleware software, such as Globus[6], Condor[7] and Legion[8]. The assumptions for the models to stand are as follows. 1) A set of N jobs {J 1, J 2,, J N } to be scheduled. 2) The proportion of jobs with high QoS requirement is denoted by P. Therefore, the number of jobs with high QoS requirement is equal to P*N, while the number of jobs with low QoS requirement is equal to (1-P)*N. 3) For each job J i with high QoS requirement, the deadline associate with it is denoted by D i. 4) Two set of hosts are defined. One is high QoS resources and the other is low QoS resources. For each set, there are M hosts {H 1, H 2,, H M } on which the jobs will be executed. 5) An array etc is a system parameter to describe the estimated time of each job to be completed on its host. The array element etc i,j indicates the estimated time to complete the ith job on the jth host. 6) Communication overhead for transferring a job from host i to host j is denoted by com i,j. 7) The expected completion time of job i on host j is defined as CT i,j. 8) The precedence constraint among the jobs is specified by connectivity, which is the probability of a job having a data dependency with a previous job. 9) Only one job can be executed on one specific host at one time. III. PROPOSED ALGORITHMS The QSF framework is designed to handle the scheduling of jobs with precedence constraints for the execution in a Grid system with QoS concerns. It schedules jobs with high QoS requirements to be executed on a resource providing high quality of service avoiding the case that the jobs requesting high QoS service waits for the jobs with low QoS requirement. The QSF scheme works in four steps. First, the scheduling problem is formalized with precedence constraints. Second, the jobs ineligible for the execution are filtered out of the scheduling. Meanwhile, the eligible jobs are divided into two sets. One is with high QoS requirement and the other is with low QoS requirement. Third, the eligible jobs including those waiting to be executed and those already in execution are scheduled according to the scheduling algorithms. And last, upon detection of a rescheduling event, the algorithm will initiate a new formalization of scheduling problem and repeat the scheduling process with new parameters. The rescheduling event is defined as the completion of one scheduled job. QSF framework aims to select the optimal match from the pool of eligible jobs and pool of available hosts. The mapping of jobs to hosts is obtained by using QSF with a simple O(M*N) algorithm. In this paper, we propose two scheduling algorithms to associate with the QSF framework. One is the Extension of Best-fit Algorithm (EBFA) and the other is Minimum Scheduling Hole First (MSHF). In the design of the two scheduling algorithms, communication overhead has been considered as an important factor for the decision making. A. QSF-EBFA Algorithm EBFA evolved from the proposal in [5], which is named as the best-fit algorithm. It has been proved to be efficient to schedule jobs in a Grid system without consideration of communication overhead. Our proposed EBFA algorithm is an extension of the best-fit algorithm by considering the communication overhead. As defined above, communication overhead for transferring data structure, due to two jobs A and B with precedence constraints, from host i to host j is denoted by com i,j. If both jobs are executed on the same host, the communication overhead will be zero, which means that no time is required to transfer the data structure. Otherwise, the communication overhead will be equal to com i,j. We present the EBFA algorithm in Fig.1, which schedules each eligible job in turn on a host, where it could be completed the earliest, taking into account the communication overhead. for (each job J i with high QoS requirements in the selected {for each host satisfying the high QoS requirements H m {if job i can be executed without waiting for data CT i,m =HAT m +etc i,m, where HAT m indicates the host available time of host m; produced from job j, which is executed on host n; if m=n, CT i,m =HAT m +etc i,m ; otherwise, CT i,m =HAT m +etc i,m + com n,m ;} select the host gives the job minimum completion time, which is denoted by m s ; for (each job J i with low QoS requirements in the selected {for each host satisfying the high and low QoS requirements H m {if job i can be executed without waiting for data CT i,m =HAT m +etc i,m ; produced from job j, which is executed on host n; if m=n, CT i,m =HAT m +etc i,m ; otherwise, CT i,m =HAT m +etc i,m + com n,m ;} select the host gives the job minimum completion time, which is denoted by m s ; B. Algorithm Fig. 1. Overview of QSF-EBFA algorithm 407

3 for (each job J i with high QoS requirements in the selected {for each host satisfying the high QoS requirements H m {if job i can be executed without waiting for data CT i,m =HAT m +etc i,m ; SH i,m =0, where SH denotes the scheduling hole produced from job j, which is executed on host n; if m=n CT i,m =HAT m +etc i,m and SH i,m =0; otherwise CT i,m =HAT m +etc i,m + com n,m and SH i,m = com n,m H m is denoted as an ineligible host for J i ;} among those eligible hosts that will not produce scheduling hole, that is, SH i,m =0 select the one gives the job minimum completion time, which is denote by m s ; for (each job J i with low QoS requirements in the selected {for each host satisfying the high and low QoS requirements H m {if job i can be executed without waiting for the CT i,m =HAT m +etc i,m ; SH i,m =0; structure X produced from job j, which is executed on host n; if m=n CT i,m =HAT m +etc i,m and SH i,m =0; otherwise CT i,m =HAT m +etc i,m + com n,m and SH i,m = com n,m ; H m is denoted as an ineligible host for J i ; among those eligible hosts that will not produce scheduling hole, that is, SH i,m =0 select the one gives the job minimum completion time, which is denote by m s ; Fig. 2. Overview of algorithm In this subsection, we propose a new scheduling algorithm, MSHF, which can largely reduce the time waste in scheduling. The logic behind the MSHF is to schedule the jobs with precedence constraints on the same host. Combined with the MSHF algorithm, the scheduling can achieve a greater improvement than the QSF-EBFA algorithm. The motivation of this proposal is to improve the performance of the QSF-EBFA algorithm. The fundamental problem in the EBFA algorithm is that it may cause a large scheduling hole on the selected host. The scheduling hole is defined as the time difference between the moment when the job starts execution and the moment when the host becomes available. For instance, if host i is selected to execute a job m, the job may be scheduled to be executed later due to the unavailability of data from job n, even though, by that time, the host i may have already become available for a while. The scheduling hole is the idle period for the host, which will waste the computing resources. The large scheduling hole in the EBFA is due to the way it selects hosts. The algorithm selects the hosts without the capability to reduce the communication overhead. This can downgrade the performance. The proposal of the MSHF algorithm can remedy this performance decay. The algorithm is described in Fig.2. The complexity of the MSHF algorithm can be evaluated based on its operations. It has one job selection procedure and one host selection procedure. The job selection procedure is to select the eligible job to be scheduled. And the host selection procedure is to select the suitable host to execute the job. The host selection algorithm simply loops with the number of the hosts, M, so that its complexity is O(M) in the worst case. Similarly, the complexity of the algorithm to select one job will be O(M) in the worst case. To schedule all the jobs, the algorithm will take O(M*N). IV. PERFORMANCE ANALYSIS We evaluate the performance of the QSF-EBFA and the algorithms by comparing them with the algorithms. The performance is measured by the makespan and tardy rate. The objectives of the simulations are twofold. First, we demonstrate the superior performance of the QSF framework by showing that the makespan and tardy rate can be reduced significantly by considering QoS factors. Second, the proposed scheduling algorithms when combined with the QSF framework are able to reduce the communication overhead greatly, thereby achieve better performance. A. Simulation Design The simulation experiments were started by a set of MATLAB functions developed to model the general behaviors of the Grid system and application models. In order to make the Grid environment heterogeneous, two sets of hosts are designed, where each host has different service time. One set of hosts are high QoS resources with 100Mb/s bandwidth and the other set of hosts are low QoS resources with 10Mb/s bandwidth. The execution of each job on each host is assumed to be a random number extracted from a Gaussian distribution with average L and variance L*H, where L denotes the average execution time for each job and H denotes the degree of heterogeneity for the Grid environment. In order to simulate the errors on the execution time estimated, three parameters, i.e., E, E 1 and E 2, are defined. For each host, E 1 is defined as a random number extracted from a Gaussian distribution with average 0 and variance E. E 2 is defined as a random number extracted from a Gaussian distribution with average E 1 and variance absolute (E 1 ). For each job s execution time estimated on this host, an error of E 2 will be introduced. Meanwhile, Jobs are randomly divided among three dependent batches. 408

4 for M ranging from 5 to 25 in the steps of 10 for L from 60s to 300s in the steps of 120s for H from 0.2 to 0.5 in the steps of 0.15 for N from 600 to 1800 in the step of 300 build up different scenarios; simulate different algorithms on different scenarios; Fig. 3. Four loops to build up simulation scenarios The communication time between two hosts among high QoS resources and the communication time between two hosts among low QoS resources are assumed to be a random number extracted from a Gaussian distribution with average L*R and (1/10)*L*R as well as variance L*R *H and (1/10)*L*R*H respectively, where R denotes the ratio of communication overhead in the set of low QoS resources to execution time of jobs. For the jobs with high QoS requirements, the time constraint on the jobs is expressed as the deadline. Deadline is assumed to be a random variable following a Gaussian distribution with average R 1 *L and variance R 1 *L*H, where R 1 denotes the ratio of average deadline to execution time of jobs. In order to evaluate the performance with respect to the number of hosts (M), the average execution time for each job (L), the degree of heterogeneity for the Grid environment (H) and the number of jobs to be scheduled (N), fours loops are used to generate different values of these four parameters so as to build up different scenarios, which is shown in Fig. 3. B. Simulation Results We produce 2 groups of simulation results. The first group of the simulation results will show the performance comparison in terms of makespan. In Fig. 4, our study focuses on the makespan versus different scenarios. In this experiment, R is set to 0.5, the proportion of jobs with high QoS requirements P is set to 50% and the error is set to 0.2. In order to have different scenarios, the algorithm shown in Fig. 3 is used. From the definition of the algorithm, we can have 135 different scenarios (N M *N L *N H *N N ) in the experimentation. From the figure, we can find that, the results are divided into three regions (1~45, 46~90, 91~135). According to the algorithm, the generation of these three regions is due to the loop of M, which is the number of hosts. It is easy to find that when M increases, makespan decreases. Meanwhile, for each distinct region, the region can be divided into three sub-regions. It is generated because of the loop of L, which is the mean execution time. When L increases, execution time increases. For each sub-region, there are three pairs of results, each of which is generated due to the loop of H, namely the degree of heterogeneity. Similarly, for each pair of result, makespan is increased with N, i.e., the number of jobs. From Fig. 4, we can draw the conclusion that supported by the QSF framework, the EBFA and the QSF algorithms significantly outperform the GSTR algorithm. This is because that the QSF is able to schedule the jobs with high QoS requirements first, which can avoid the case that the job with Time (s) Time (s) 8 x * QSF-EBFA + o Scenarios 4.5 x QSF-EBFA Fig. 4. Makespan versus scenarios Proportion of the tasks with high QoS requirements (%) Fig. 5. Makespan versus proportion of the jobs with high QoS requirements low QoS requirements occupy the high QoS resources while jobs with high QoS requirements have to wait even if the low QoS resources are idle. On the other hand, it is shown in Fig. 4 that the performance of the MSHF algorithm is better than that of the EBFA algorithm when combining with the QSF framework. It is predictable since MSHF is able to reduce the scheduling holes. Fig. 5 illustrates the makespan versus the proportion of jobs with high QoS requirements. For example, proportion of 70% means 70% jobs request high QoS resources, while 30% jobs have no QoS requirements. In this experiment, M is set to 15, L is set to 300s, R is set to 0.5, error is set to 0.2 and N is set to As shown in the figure, the QSF framework works better than the GSTR scheme in terms of makespan, especially when half of jobs have high QoS requirements. When the proportion is increased to 1, all the jobs request high QoS resources. In this case, the QSF and the GSTR will get the same performance because no low QoS jobs will compete for the resources. Among the three algorithms, the algorithm can achieve the best performance. It is due to the fact that the 409

5 8 x QSF-EBFA 7.5 QSF- EBFA 50 Time (s) Tardy rate (%) Ratio of communication overhead to execution time Fig. 6. Makespan versus ratio of communication overhead to execution time Proportion of the jobs with high QoS requirements (%) Fig. 8. Tardy rate versus proportion of the jobs with high QoS requirements QSF-EBFA Tardy rate (%) Tardy rate (%) * QSF-EBFA + o Scenarios Fig. 7. Tardy rate versus scenarios MSHF scheme is capable to reduce the scheduling holes, which can increase the utilization of the computing resources. Fig. 6 shows the makespan versus R, i.e, the ratio of communication overhead in the low QoS set of resources to execution time of jobs. In this experiment, M is set to to 5, L is set to 300s, N is set to 1800, H is set to 0.5, error is set to 0.2, and the proportion of jobs with high QoS requirements is set to 50%. It is clear that the MSHF algorithm has achieved a significant improvement in terms of makespan, particularly, when the ratio is large. This is because the cost of data communication in a Grid system can significantly affect the scheduling and execution of jobs, especially when the communication overhead is comparable to the execution time. The MSHF algorithm aims to reduce the scheduling holes on the hosts. Meanwhile, the algorithm presents stable results when the ratio is increased. It is expected since communication overhead is reduced in the MSHF algorithm and it won t affect the performance of the scheduling. Similarly, it is shown that the QSF scheme works better than the GSTR scheme. The second group of the simulation results will show the performance comparison in terms of trady rate. Figure 8 presents tardy rate versus different scenarios. In this set of experiment, we assume that R is 0.5, R 1 is 10, the proportion of jobs with high QoS requirements P is 50% and the error is set to Ratio of average deadline to execution time (%) Fig. 9. Tardy rate versus ratio of average deadline to execution time 0.2. Different scenarios are built up by using the algorithm shown in Fig. 3. As shown in the figure, there are 135 different scenarios. Similar as Figure 5, the results are divided into three regions (1~45, 46~90, 91~135), which are generated by the loop of M. The regions are further divided into three sub-regions, which are generated because of the loop of L. There are three pairs of results in each sub-region, which is generated due to the loop of H. And for each pair of results, the tardy rate is increased, which is because of the loop of N. As Fig. 7 shows, the QSF-EBFA and the algorithms outperform the algorithm significantly. This is because that the QSF framework has the ability to schedule jobs by taking QoS factors into the account. However, the GSTR scheme schedules jobs with high or low QoS requirements with the same priority. Therefore, the jobs with high QoS requirements may be blocked by those with low QoS requirement leading to high tardy rate. Fig. 8 compares the characteristic of tardy rate under varying the proportion of jobs with high QoS requirements. In this experiment, M is set to to 15, L is set to 300s, R is set to 0.5, R 1 is 40, error is set to 0.2, D is set to 3000s and N is set to As shown in the figure, the tardy rate of three algorithms is increased with the proportion of jobs with high QoS requirements. It is clear that QSF outperforms GSTR in terms of tardy rate. This is because the QSF algorithms can arrange 410

6 the jobs with high QoS requirements to be scheduled faster than the jobs with low QoS requirements, therefore resulting in low tardy rate. Also, the algorithm is better than QSF-EBFA. There is because MSHF algorithm is able to reduce the scheduling holes, leading to better performance. Fig. 9 shows the relationship between the tardy rate and average deadline. We assume that M is equal to 5, L is set to 300s, R is set to 0.5, error is set to 0.2, P is set to 50% and N is set to 600.We can clearly see that the tardy rate of these three algorithms decreases when the average deadline increases. As the figure shows, the QSF scheme with associated algorithms has achieved a significant improvement in terms of tardy rate. This is because the QSF scheme is capable to take the QoS factors as important factors in the scheduling. V. CONCLUSION In this paper, we introduce a novel scheduler QSF, which is designed to handle jobs with high QoS requirements as well as those with low QoS requirements in the Grid computing environments. With the QSF scheme, the makespan and tardy rate can be reduced significantly. Meanwhile, a novel scheduling algorithm, namely MSHF, is proposed to associate with the QSF scheme. The MSHF algorithm is to reduce the scheduling holes by taking the communication overhead into account. The small scheduling hole can lead to lower the performance in terms of increasing the makespan. Numerical studies have been conducted by extensive simulations. As the results show, the QSF framework can reduce the makespan and tardy rate significantly. In addition, the performance with respect to the proportion of high QoS jobs as well as the ratio of communication overhead to execution time/deadline is evaluated. The studies have shown that the scheduling can achieve the best performance because it has to ability to take QoS factors into account and to reduce the scheduling holes simultaneously. REFERENCES [1] I. Foster, The Grid: A New Infrastructure for 21 st Century Science, Physics Today, 55 (2) 2002, pp [2] D. F. Baca, Allocation modules to processors in a distributed system, IEEE Transactions on Software Engineering, Vol. 15, No. 11, pp , [3] C. A. Bohn and G. B. Lamont, Load balancing for heterogeneous clusters of PCs, Future Generation Computer Systems, Vol. 18, pp , [4] B. R. Carter, D. W. Watson, F. R. F. Freund, K. Elaine, M. Francesca and H. J. Siegel, Generational scheduling for dynamic job management in heterogeneous computing systems, Journal of Information Sciences, Vol. 106, pp , [5] O. Lucchese, F.; Huerta Yero, EJ; Sambatti, FS, An adaptive scheduler for Grids, Journal of Grid Computing, Vol. 4, No. 1. pp. 1-17, [6] I. Foster, C. Kesselman, J. Nick and S. Tuecke, Grid services for distributed system integration, Computer, vol. 35, No. 6, [7] J. H. Epenema, M. Livny, R. Van Dantzing, X. Evers and J. Pruyne, A worldwide flock of condors: Load sharing among workstations clusters, Future Generation Computer Systems, Vol. 12, pp , [8] A. S. Grimshaw and W. A. Wulf, Legion A view from 50,000 feet Los Alamitos, California, in Proceedings of the Fifth IEEE International Symposium on High Performance Distributed Computing, Agosto, IEEE Computer Society Press,

LOAD SHARING IN HETEROGENEOUS DISTRIBUTED SYSTEMS

LOAD SHARING IN HETEROGENEOUS DISTRIBUTED SYSTEMS Proceedings of the 2 Winter Simulation Conference E. Yücesan, C.-H. Chen, J. L. Snowdon, and J. M. Charnes, eds. LOAD SHARING IN HETEROGENEOUS DISTRIBUTED SYSTEMS Helen D. Karatza Department of Informatics

More information

Gang Scheduling Performance on a Cluster of Non-Dedicated Workstations

Gang Scheduling Performance on a Cluster of Non-Dedicated Workstations Gang Scheduling Performance on a Cluster of Non-Dedicated Workstations Helen D. Karatza Department of Informatics Aristotle University of Thessaloniki 54006 Thessaloniki, Greece karatza@csd.auth.gr Abstract

More information

SEQUENCING & SCHEDULING

SEQUENCING & SCHEDULING SEQUENCING & SCHEDULING November 14, 2010 1 Introduction Sequencing is the process of scheduling jobs on machines in such a way so as to minimize the overall time, cost and resource usage thereby maximizing

More information

TBS: A Threshold Based Scheduling in Grid Environment

TBS: A Threshold Based Scheduling in Grid Environment TBS: A Threshold Based Scheduling in Grid Environment Sanjaya Kumar Panda Department of Computer Science and Engineering National Institute of Technology Rourkela, India sanjayauce@gmail.com Pabitra Mohan

More information

Preemptive Resource Management for Dynamically Arriving Tasks in an Oversubscribed Heterogeneous Computing System

Preemptive Resource Management for Dynamically Arriving Tasks in an Oversubscribed Heterogeneous Computing System 2017 IEEE International Parallel and Distributed Processing Symposium Workshops Preemptive Resource Management for Dynamically Arriving Tasks in an Oversubscribed Heterogeneous Computing System Dylan Machovec

More information

Scheduling and Resource Management in Grids

Scheduling and Resource Management in Grids Scheduling and Resource Management in Grids ASCI Course A14: Advanced Grid Programming Models Ozan Sonmez and Dick Epema may 13, 2009 1 Outline Resource Management Introduction A framework for resource

More information

INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS

INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS ISSN 2320-7345 AN IMPRESSION ON PERFORMANCE METRICS FOR SCHEDULING PROBLEM IN GRID COMPUTING ENVIRONMENT Abstract D. Thilagavathi

More information

A Machine Setup Model for TFT-LCD Cell Back-End Process

A Machine Setup Model for TFT-LCD Cell Back-End Process A Machine Setup Model for TFT-LCD Cell Back-End Process Y.-C. Chang 1, P.-S. Chen 2, P.-C. Chen 1 1 Department of Industrial Engineering and Management, National Chiao Tung University, Hsinchu, Taiwan

More information

Hybrid search method for integrated scheduling problem of container-handling systems

Hybrid search method for integrated scheduling problem of container-handling systems Hybrid search method for integrated scheduling problem of container-handling systems Feifei Cui School of Computer Science and Engineering, Southeast University, Nanjing, P. R. China Jatinder N. D. Gupta

More information

A Proactive Fault Tolerant Strategy for Desktop Grid

A Proactive Fault Tolerant Strategy for Desktop Grid A Tolerant Strategy for Desktop Grid Geeta Arora 1,Dr. Shaveta Rani 2,Dr. Paramjit Singh 3 1 Research Scholar, Dept. of Computer Applications, I.K.Gujral Punjab Technical University, Kapurthala,India 2,

More information

Resource Allocation Strategies in a 2-level Hierarchical Grid System

Resource Allocation Strategies in a 2-level Hierarchical Grid System st Annual Simulation Symposium Resource Allocation Strategies in a -level Hierarchical Grid System Stylianos Zikos and Helen D. Karatza Department of Informatics Aristotle University of Thessaloniki 5

More information

Presenting a New Ant Colony Optimization Algorithm (ACO) for Efficient Job Scheduling in Grid Environment

Presenting a New Ant Colony Optimization Algorithm (ACO) for Efficient Job Scheduling in Grid Environment Presenting a New Ant Colony Optimization Algorithm (ACO) for Efficient Job Scheduling in Grid Environment Firoozeh Ghazipour Science and Research Branch Islamic Azad University, Kish, Iran Seyyed Javad

More information

Integration of Process Planning and Scheduling Functions for Batch Manufacturing

Integration of Process Planning and Scheduling Functions for Batch Manufacturing Integration of Process Planning and Scheduling Functions for Batch Manufacturing A.N. Saravanan, Y.F. Zhang and J.Y.H. Fuh Department of Mechanical & Production Engineering, National University of Singapore,

More information

Job co-allocation strategies for multiple high performance computing clusters

Job co-allocation strategies for multiple high performance computing clusters Cluster Comput (2009) 12: 323 340 DOI 10.1007/s10586-009-0087-x Job co-allocation strategies for multiple high performance computing clusters Jinhui Qin Michael A. Bauer Received: 28 April 2008 / Accepted:

More information

Decentralized Scheduling of Bursty Workload on Computing Grids

Decentralized Scheduling of Bursty Workload on Computing Grids Decentralized Scheduling of Bursty Workload on Computing Grids Juemin Zhang, Ningfang Mi, Jianzhe Tai and Waleed Meleis Department of Electrical and Computer Engineering Northeastern University, Boston,

More information

MULTI OBJECTIVE BEE COLONY OPTIMIZATION FRAMEWORK FOR GRID JOB SCHEDULING

MULTI OBJECTIVE BEE COLONY OPTIMIZATION FRAMEWORK FOR GRID JOB SCHEDULING MULTI OBJECTIVE BEE COLONY OPTIMIZATION FRAMEWORK FOR GRID JOB SCHEDULING Sana Alyaseri 1 and Ku Ruhana Ku-Mahamud 2 1 Nazwa College of Technology, Oman, sana.alyaseri@nct.edu.om 2 Universiti Utara Malaysia,

More information

Paper 30 Centralized versus Market-based Task Allocation in the Presence of Uncertainty

Paper 30 Centralized versus Market-based Task Allocation in the Presence of Uncertainty Paper 30 Centralized versus Market-based Task Allocation in the Presence of Uncertainty Abstract While there have been some efforts to compare centralized versus market based approaches to general task

More information

Minimizing Makespan for Machine Scheduling and Worker Assignment Problem in Identical Parallel Machine Models Using GA

Minimizing Makespan for Machine Scheduling and Worker Assignment Problem in Identical Parallel Machine Models Using GA , June 30 - July 2, 2010, London, U.K. Minimizing Makespan for Machine Scheduling and Worker Assignment Problem in Identical Parallel Machine Models Using GA Imran Ali Chaudhry, Sultan Mahmood and Riaz

More information

CHAPTER 6 DYNAMIC SERVICE LEVEL AGREEMENT FOR GRID RESOURCE ALLOCATION

CHAPTER 6 DYNAMIC SERVICE LEVEL AGREEMENT FOR GRID RESOURCE ALLOCATION 158 CHAPTER 6 DYNAMIC SERVICE LEVEL AGREEMENT FOR GRID RESOURCE ALLOCATION 6.1 INTRODUCTION In a dynamic and heterogeneous Grid environment providing guaranteed quality of service for user s job is fundamentally

More information

Resource Scheduling in Hybrid Grid Environment

Resource Scheduling in Hybrid Grid Environment Resource Scheduling in Hybrid Grid Environment Dr. N. Malarvizhi Professor & Head, Department of Information Technology Jawahar Engineering College Chennai, India nmv_94@yahoo.com Dr. N. Sankar Ram Professor

More information

Cloud Load Balancing Based on ACO Algorithm

Cloud Load Balancing Based on ACO Algorithm Cloud Load Balancing Based on ACO Algorithm Avtar Singh, Kamlesh Dutta, Himanshu Gupta Department of Computer Science, National Institute of Technology,Hamirpur, HP, India Department of Computer Science,

More information

Enhancing Resource Reservation by Applying Defer Time in FCFS Scheme for Grid Computing Environments

Enhancing Resource Reservation by Applying Defer Time in FCFS Scheme for Grid Computing Environments 2017 IJSRST Volume 3 Issue 3 Print ISSN: 2395-6011 Online ISSN: 2395-602X Themed Section: Science and Technology Enhancing Reservation by Applying Defer Time in FCFS Scheme for Grid Computing Environments

More information

Energy-Efficient Scheduling of Interactive Services on Heterogeneous Multicore Processors

Energy-Efficient Scheduling of Interactive Services on Heterogeneous Multicore Processors Energy-Efficient Scheduling of Interactive Services on Heterogeneous Multicore Processors Shaolei Ren, Yuxiong He, Sameh Elnikety University of California, Los Angeles, CA Microsoft Research, Redmond,

More information

A HYBRID GENETIC ALGORITHM FOR JOB SHOP SCHEUDULING

A HYBRID GENETIC ALGORITHM FOR JOB SHOP SCHEUDULING A HYBRID GENETIC ALGORITHM FOR JOB SHOP SCHEUDULING PROF. SARVADE KISHORI D. Computer Science and Engineering,SVERI S College Of Engineering Pandharpur,Pandharpur,India KALSHETTY Y.R. Assistant Professor

More information

Ph.D. Defense: Resource Allocation Optimization in the Smart Grid and High-performance Computing Tim Hansen

Ph.D. Defense: Resource Allocation Optimization in the Smart Grid and High-performance Computing Tim Hansen Ph.D. Defense: Resource Allocation Optimization in the Smart Grid and High-performance Computing Tim Hansen Department of Electrical and Computer Engineering Colorado State University Fort Collins, Colorado,

More information

GROUPING BASED USER DEMAND AWARE JOB SCHEDULING APPROACH FOR COMPUTATIONAL GRID

GROUPING BASED USER DEMAND AWARE JOB SCHEDULING APPROACH FOR COMPUTATIONAL GRID GROUPING BASED USER DEMAND AWARE JOB SCHEDULING APPROACH FOR COMPUTATIONAL GRID P.SURESH Assistant Professor (Senior Grade), Department of IT, Kongu Engineering College, Perundurai, Erode Tamilnadu, India

More information

CROWNBench: A Grid Performance Testing System Using Customizable Synthetic Workload

CROWNBench: A Grid Performance Testing System Using Customizable Synthetic Workload CROWNBench: A Grid Performance Testing System Using Customizable Synthetic Workload Xing Yang, Xiang Li, Yipeng Ji, and Mo Sha School of Computer Science, Beihang University, Beijing, China {yangxing,

More information

A Particle Swarm Optimization Approach for Workflow Scheduling on Cloud Resources Priced by CPU Frequency

A Particle Swarm Optimization Approach for Workflow Scheduling on Cloud Resources Priced by CPU Frequency A Particle Swarm Optimization Approach for Workflow Scheduling on Cloud Resources Priced by CPU Frequency Thiago A. L. Genez, Ilia Pietri, Rizos Sakellariou, Luiz F. Bittencourt and Edmundo R. M. Madeira

More information

Graph Optimization Algorithms for Sun Grid Engine. Lev Markov

Graph Optimization Algorithms for Sun Grid Engine. Lev Markov Graph Optimization Algorithms for Sun Grid Engine Lev Markov Sun Grid Engine SGE management software that optimizes utilization of software and hardware resources in heterogeneous networked environment.

More information

Slack Assessment of the Real Time Scheduling Algorithms

Slack Assessment of the Real Time Scheduling Algorithms International Journal of Control Theory and Applications ISSN : 0974-5572 International Science Press Volume 9 Number 43 2016 Slack Assessment of the Real Time Scheduling Algorithms P. Ramesh a and Uppu

More information

Trust Based Grid Scheduling Algorithm for Commercial Grids

Trust Based Grid Scheduling Algorithm for Commercial Grids International Conference on Computational Intelligence and Multimedia Applications 2007 Trust Based Grid Scheduling Algorithm for Commercial Grids S.ThamaraiSelvi, P.Balakrishnan, R.Kumar, K.Rajendar Department

More information

Design and Implementation of LST Based Dynamic Scheduler on Real Time OS

Design and Implementation of LST Based Dynamic Scheduler on Real Time OS Design and Implementation of LST Based Dynamic Scheduler on Real Time OS Prasad M Kamath, Rakesh Belagali, Sushant S Kulkarni, Vinayak Hegde, Geetishree Mishra Department of Electronics and Communication,

More information

A Modeling Tool to Minimize the Expected Waiting Time of Call Center s Customers with Optimized Utilization of Resources

A Modeling Tool to Minimize the Expected Waiting Time of Call Center s Customers with Optimized Utilization of Resources A Modeling Tool to Minimize the Expected Waiting Time of Call Center s Customers with Optimized Utilization of Resources Mohsin Iftikhar Computer Science Department College of Computer and Information

More information

PARALLEL LINE AND MACHINE JOB SCHEDULING USING GENETIC ALGORITHM

PARALLEL LINE AND MACHINE JOB SCHEDULING USING GENETIC ALGORITHM PARALLEL LINE AND MACHINE JOB SCHEDULING USING GENETIC ALGORITHM Dr.V.Selvi Assistant Professor, Department of Computer Science Mother Teresa women s University Kodaikanal. Tamilnadu,India. Abstract -

More information

Enhanced Algorithms for Multi-Site Scheduling

Enhanced Algorithms for Multi-Site Scheduling Enhanced Algorithms for Multi-Site Scheduling Carsten Ernemann 1, Volker Hamscher 1, Ramin Yahyapour 1, and Achim Streit 2 1 Computer Engineering Institute, University of Dortmund, 44221 Dortmund, Germany,

More information

INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS ISSN

INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS ISSN INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS ISSN 2320-7345 SURVEY ON META-SCHEDULER IN GRID ENVIRONMENT P.Priscilla 1, Mr. P.Karthikeyan 2 1 PG Full Time Student, CSE Department,

More information

An Evaluation of Economy-based Resource Trading and Scheduling on Computational Power Grids for Parameter Sweep Applications

An Evaluation of Economy-based Resource Trading and Scheduling on Computational Power Grids for Parameter Sweep Applications An Evaluation of Economy-based Resource Trading and Scheduling on Computational Power Grids for Parameter Sweep Applications Rajkumar Buyya, Jonathan Giddy, and David Abramson School of Computer Science

More information

AN ADVANCED IWD BASED HYPER-HEURISTIC WORKFLOW SCHEDULING IN COMPUTATIONAL GRID

AN ADVANCED IWD BASED HYPER-HEURISTIC WORKFLOW SCHEDULING IN COMPUTATIONAL GRID AN ADVANCED IWD BASED HYPER-HEURISTIC WORKFLOW SCHEDULING IN COMPUTATIONAL GRID S. Gokuldev 1, C. Sowntharya 2 and S. Manishankar 1 1 Department of Computer Science, Amrita VishwaVidyapeetham, Mysore Campus,

More information

Optimized Assignment of Independent Task for Improving Resources Performance in Computational Grid

Optimized Assignment of Independent Task for Improving Resources Performance in Computational Grid Optimized Assignment of Independent Task for Improving Resources Performance in Computational Grid Sarpreet Singh 1, R.K. Bawa 2, 1 Assistant Professor, Department of Computer Science& Engineering, Sri

More information

Application of Min-Min and Max-Min Algorithm for Task Scheduling in Cloud Environment Under Time Shared and Space Shared VM Models *

Application of Min-Min and Max-Min Algorithm for Task Scheduling in Cloud Environment Under Time Shared and Space Shared VM Models * International Journal of Computing Academic Research (IJCAR) ISSN 235-9184, Volume 4, Number 6 (December 215), pp.182-19 MEACSE Publications http://www.meacse.org/ijcar Application of Min-Min and Max-Min

More information

Characterization of Backfilling Strategies for Parallel Job Scheduling

Characterization of Backfilling Strategies for Parallel Job Scheduling Characterization of Backfilling Strategies for Parallel Job Scheduling Srividya Srinivasan Rajkumar Kettimuthu Vijay Subramani P. Sadayappan Department of Computer and Information Science The Ohio State

More information

Availability Modeling of Grid Computing Environments Using SANs

Availability Modeling of Grid Computing Environments Using SANs Availability Modeling of Grid Computing Environments Using SANs Reza Entezari-Maleki Department of Computer Engineering, Sharif University of Technology, Tehran, Iran E-mail: entezari@ce.sharif.edu Ali

More information

Scheduling Dynamic Parallel Loop Workflow in Cloud Environment

Scheduling Dynamic Parallel Loop Workflow in Cloud Environment Information Technology Scheduling Dynamic Parallel Loop Workflow in Cloud Environment Sucha SMANCHAT * and Kanchana VIRIYAPANT Faculty of Information Technology, King Mongkut s University of Technology

More information

A New Algorithm for Scheduling Parallel Tasks in Utility Grids using Reserve Resources for Tasks with Deadline

A New Algorithm for Scheduling Parallel Tasks in Utility Grids using Reserve Resources for Tasks with Deadline A New Algorithm for Scheduling Parallel Tasks in Utility Grids using Reserve Resources for Tasks with Deadline Leyli Mohammad khanli #, Motahareh Ahmadi * # Computer Department, Tabriz University Tabriz

More information

APPENDIX 1 Processing time (pij) Release date (rj) Due date (dj): Weight (wj)

APPENDIX 1 Processing time (pij) Release date (rj) Due date (dj): Weight (wj) 172 APPENDIX 1 In all the scheduling problems considered the number of jobs and the number of machines are assumed to be finite. The number of jobs is denoted by N and the number of machines by M. usually,

More information

Computer Cluster Scheduling Algorithm Based on Time Bounded Dynamic Programming

Computer Cluster Scheduling Algorithm Based on Time Bounded Dynamic Programming Computer Cluster Scheduling Algorithm Based on Time Bounded Dynamic Programming I.Grudenic and N. Bogunovic Department of Electronics, Microelectronics, Computer and Intelligent Systems Faculty of electrical

More information

Exploiting Dynamic Workload Variation in Low Energy Preemptive Task Scheduling

Exploiting Dynamic Workload Variation in Low Energy Preemptive Task Scheduling Exploiting Dynamic Workload Variation in Low Energy Preemptive Scheduling Xiaobo Sharon Department of Electrical and Electronic Engineering Department of Computer Science and Engineering Hong Kong University

More information

A HYBRID ALGORITHM TO MINIMIZE THE NUMBER OF TARDY JOBS IN SINGLE MACHINE SCHEDULING

A HYBRID ALGORITHM TO MINIMIZE THE NUMBER OF TARDY JOBS IN SINGLE MACHINE SCHEDULING DAAAM INTERNATIONAL SCIENTIFIC BOOK 2010 pp. 549-558 CHAPTER 48 A HYBRID ALGORITHM TO MINIMIZE THE NUMBER OF TARDY JOBS IN SINGLE MACHINE SCHEDULING BANCILA, D.; BUZATU, C. & FOTA, A. Abstract: Starting

More information

Integrating New Cost Model into HMA-Based Grid Resource Scheduling

Integrating New Cost Model into HMA-Based Grid Resource Scheduling Integrating New Cost Model into HMA-Based Grid Resource Scheduling Jun-yan Zhang, Fan Min, and Guo-wei Yang College of Computer Science and Engineering, University of Electronic Science and Technology

More information

A Novel Workload Allocation Strategy for Batch Jobs

A Novel Workload Allocation Strategy for Batch Jobs Int. J. Com. Net. Tech. 1, No. 1, 1-17 (2013) 1 International Journal of Computing and Network Technology An International Journal A Novel Workload Allocation Strategy for Batch Jobs A. Shenfield 1 and

More information

Task Assignment Problem of Robots in a Smart Warehouse Environment. Zhenping Li, Wenyu Li, Lulu Jiang. Beijing Wuzi University, Beijing, China

Task Assignment Problem of Robots in a Smart Warehouse Environment. Zhenping Li, Wenyu Li, Lulu Jiang. Beijing Wuzi University, Beijing, China Management Studies, July-Aug. 2016, Vol. 4, No. 4, 167-175 doi: 10.17265/2328-2185/2016.04.004 D DAVID PUBLISHING Task Assignment Prolem of Roots in a Smart Warehouse Environment Zhenping Li, Wenyu Li,

More information

Incentive-Based P2P Scheduling in Grid Computing

Incentive-Based P2P Scheduling in Grid Computing Incentive-Based P2P Scheduling in Grid Computing Yanmin Zhu 1, Lijuan Xiao 2, Lionel M. Ni 1, and Zhiwei Xu 2 1 Department of Computer Science Hong Kong University of Science and Technology Clearwater

More information

SINGLE MACHINE SEQUENCING. ISE480 Sequencing and Scheduling Fall semestre

SINGLE MACHINE SEQUENCING. ISE480 Sequencing and Scheduling Fall semestre SINGLE MACHINE SEQUENCING 2011 2012 Fall semestre INTRODUCTION The pure sequencing problem is a specialized scheduling problem in which an ordering of the jobs completely determines a schedule. Moreover,

More information

An Optimal Service Ordering for a World Wide Web Server

An Optimal Service Ordering for a World Wide Web Server An Optimal Service Ordering for a World Wide Web Server Amy Csizmar Dalal Hewlett-Packard Laboratories amy dalal@hpcom Scott Jordan University of California at Irvine sjordan@uciedu Abstract We consider

More information

Packet Scheduling in Cloud by Employing Genetic Algorithm

Packet Scheduling in Cloud by Employing Genetic Algorithm Packet Scheduling in Cloud by Employing Genetic Algorithm S.Prabhu Assistant professor Department of Computer Science and Engineering Nandha Engineeirng College, Erode, Tamil Nadu, India Dr.N.Sengottaiyan

More information

Grid Computing Scheduling Jobs Based on Priority Using Backfilling

Grid Computing Scheduling Jobs Based on Priority Using Backfilling Grid Computing Scheduling Jobs Based on Priority Using Backfilling Sandip Fakira Lokhande 1, Sachin D. Chavhan 2, Prof. S. R. Jadhao 3 1-2 ME Scholar, Department of CSE, Babasaheb Naik College of Engineering,

More information

COST ESTIMATION OF ADVANCE RESERVATIONS OVER QUEUED JOBS: A QUANTITATIVE STUDY

COST ESTIMATION OF ADVANCE RESERVATIONS OVER QUEUED JOBS: A QUANTITATIVE STUDY International Journal of Modeling, Simulation, and Scientific Computing Vol. 1, No. 3 (2010) 317 332 c World Scientific Publishing Company DOI: 10.1142/S1793962310000249 COST ESTIMATION OF ADVANCE RESERVATIONS

More information

Service Proveider s Optimal Pricing for PVC and SVC Service

Service Proveider s Optimal Pricing for PVC and SVC Service The University of Kansas Technical Report Service Proveider s Optimal Pricing for PVC and SVC Service Yuhong Liu and David W. Petr ITTC-FY2000-TR-18836-02 December 1999 Project Sponsor: Sprint Corporation

More information

JOB SHOP SCHEDULING TO MINIMIZE WORK-IN-PROCESS, EARLINESS AND TARDINESS COSTS ZHU ZHECHENG A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

JOB SHOP SCHEDULING TO MINIMIZE WORK-IN-PROCESS, EARLINESS AND TARDINESS COSTS ZHU ZHECHENG A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY JOB SHOP SCHEDULING TO MINIMIZE WORK-IN-PROCESS, EARLINESS AND TARDINESS COSTS ZHU ZHECHENG A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF INDUSTRIAL AND SYSTEMS ENGINEERING NATIONAL

More information

A NEW SELECTION ALGORITHM BASED ON REPUTATION AND CONTRACT-MANAGEMENT

A NEW SELECTION ALGORITHM BASED ON REPUTATION AND CONTRACT-MANAGEMENT A NEW SELECTION ALGORITHM BASED ON REPUTATION AND CONTRACT-MANAGEMENT 1 LINGYU XU, 2 YINGHUA SUN 1 Lecturer, International College, Qingdao University,China 2 Assoc. Prof., Information and Engineering

More information

Research Statement. Nilabja Roy. August 18, 2010

Research Statement. Nilabja Roy. August 18, 2010 Research Statement Nilabja Roy August 18, 2010 1 Doctoral Dissertation Large-scale, component-based distributed systems form the backbone of many service-oriented applications, ranging from Internet portals

More information

Fixed vs. Self-Adaptive Crossover-First Differential Evolution

Fixed vs. Self-Adaptive Crossover-First Differential Evolution Applied Mathematical Sciences, Vol. 10, 2016, no. 32, 1603-1610 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2016.6377 Fixed vs. Self-Adaptive Crossover-First Differential Evolution Jason

More information

An Economic Allocation of Resources for Divisible Workloads in Grid Computing Paradigm

An Economic Allocation of Resources for Divisible Workloads in Grid Computing Paradigm European Journal of Scientific Research ISS 1450-216X Vol.65 o.3 (2011), pp. 434-443 EuroJournals Publishing, Inc. 2011 http://www.europeanjournalofscientificresearch.com An Economic Allocation of Resources

More information

Australian Journal of Basic and Applied Sciences. LB Scheduling for Advanced Reservation and Queuing Using TBRA in Grid Computing Environments

Australian Journal of Basic and Applied Sciences. LB Scheduling for Advanced Reservation and Queuing Using TBRA in Grid Computing Environments AENSI Journals Australian Journal of Basic and Applied Sciences ISSN:1991-8178 Journal home page: www.ajbasweb.com LB Scheduling for Advanced Reservation and Queuing Using TBRA in Grid Computing Environments

More information

EFFECT OF CROSS OVER OPERATOR IN GENETIC ALGORITHMS ON ANTICIPATORY SCHEDULING

EFFECT OF CROSS OVER OPERATOR IN GENETIC ALGORITHMS ON ANTICIPATORY SCHEDULING 24th International Symposium on on Automation & Robotics in in Construction (ISARC 2007) Construction Automation Group, I.I.T. Madras EFFECT OF CROSS OVER OPERATOR IN GENETIC ALGORITHMS ON ANTICIPATORY

More information

1. For s, a, initialize Q ( s,

1. For s, a, initialize Q ( s, Proceedings of the 2006 Winter Simulation Conference L. F. Perrone, F. P. Wieland, J. Liu, B. G. Lawson, D. M. Nicol, and R. M. Fujimoto, eds. A REINFORCEMENT LEARNING ALGORITHM TO MINIMIZE THE MEAN TARDINESS

More information

Analysis and Modelling of Flexible Manufacturing System

Analysis and Modelling of Flexible Manufacturing System Analysis and Modelling of Flexible Manufacturing System Swetapadma Mishra 1, Biswabihari Rath 2, Aravind Tripathy 3 1,2,3Gandhi Institute For Technology,Bhubaneswar, Odisha, India --------------------------------------------------------------------***----------------------------------------------------------------------

More information

Dynamic Matching and Scheduling of a Class of Independent Tasks onto Heterogeneous Computing Systems

Dynamic Matching and Scheduling of a Class of Independent Tasks onto Heterogeneous Computing Systems Dynamic Matching and Scheduling of a Class of Independent Tasks onto Heterogeneous Computing Systems Muthucumaru Maheswaran, Shoukat Ali, Howard Jay Siegel, Debra Hensgen,andRichardF.Freund? Department

More information

Alea Grid Scheduling Simulation Environment

Alea Grid Scheduling Simulation Environment Alea Grid Scheduling Simulation Environment Dalibor Klusáček, Luděk Matyska, and Hana Rudová Faculty of Informatics, Masaryk University, Botanická 68a, 602 00 Brno, Czech Republic {xklusac,ludek,hanka}@fi.muni.cz

More information

Economic Scheduling in Grid Computing

Economic Scheduling in Grid Computing Economic Scheduling in Grid Computing Carsten Ernemann, Volker Hamscher, and Ramin Yahyapour Computer Engineering Institute, University of Dortmund, 44221 Dortmund, Germany, (email: {carsten.ernemann,volker.hamscher,ramin.yahyapour}@udo.edu)

More information

ELC 4438: Embedded System Design Real-Time Scheduling

ELC 4438: Embedded System Design Real-Time Scheduling ELC 4438: Embedded System Design Real-Time Scheduling Liang Dong Electrical and Computer Engineering Baylor University Scheduler and Schedule Jobs are scheduled and allocated resources according to the

More information

COLLABORATIVE WORKSPACE OVER SERVICE-ORIENTED GRID

COLLABORATIVE WORKSPACE OVER SERVICE-ORIENTED GRID COLLABORATIVE WORKSPACE OVER SERVICE-ORIENTED GRID SHEN ZHIQI Information Communication Institute of Singapore, School of Electrical and Electronic Engineering, Nanyang Technological University, 50 Nanyang

More information

On the Impact of Reservations from the Grid on Planning-Based Resource Management

On the Impact of Reservations from the Grid on Planning-Based Resource Management On the Impact of Reservations from the Grid on Planning-Based Resource Management Felix Heine 1, Matthias Hovestadt 1, Odej Kao 1, and Achim Streit 2 1 Paderborn Center for Parallel Computing (PC 2 ),

More information

Parallel Cloud Computing Billing Model For Maximizing User s Utility and Provider s Cost-Efficiency

Parallel Cloud Computing Billing Model For Maximizing User s Utility and Provider s Cost-Efficiency Parallel Cloud Computing Billing Model For Maximizing User s Utility and Provider s Cost-Efficiency ABSTRACT Presented cloud computing billing model is designed for maximizing value-adding throughput of

More information

Managing Risk of Inaccurate Runtime Estimates for Deadline Constrained Job Admission Control in Clusters

Managing Risk of Inaccurate Runtime Estimates for Deadline Constrained Job Admission Control in Clusters Managing Risk of Inaccurate Runtime Estimates for Deadline Constrained Job Admission Control in Clusters Chee Shin Yeo and Rajkumar Buyya Grid Computing and Distributed Systems Laboratory Department of

More information

Hybrid performance-based workload management for multiclusters and grids

Hybrid performance-based workload management for multiclusters and grids Hybrid performance-based workload management for multiclusters and grids L. He, S.A. Jarvis, D.P. Spooner, X. Chen and G.R. Nudd Abstract: The paper addresses the dynamic scheduling of parallel jobs with

More information

A Simulation-based Multi-level Redundancy Allocation for a Multi-level System

A Simulation-based Multi-level Redundancy Allocation for a Multi-level System International Journal of Performability Engineering Vol., No. 4, July 205, pp. 357-367. RAMS Consultants Printed in India A Simulation-based Multi-level Redundancy Allocation for a Multi-level System YOUNG

More information

Rehandling Strategies for Container Retrieval

Rehandling Strategies for Container Retrieval Rehandling Strategies for Container Retrieval Tonguç Ünlüyurt and Cenk Aydin Sabanci University, Faculty of Engineering and Natural Sciences e-mail: tonguc@sabanciuniv.edu 1 Introduction In this work,

More information

An Improved Job Scheduling Algorithm Based on the First Distribution and Redistribution Strategy for Grid Computing

An Improved Job Scheduling Algorithm Based on the First Distribution and Redistribution Strategy for Grid Computing Algorithm Based on the First Distribution and Redistribution Strategy for Grid Computing 1 2 Department of Electronics and Information Engineering Hohai University Wentian College, Maanshan 243031 Anhui

More information

Dynamic Vehicle Routing and Dispatching

Dynamic Vehicle Routing and Dispatching Dynamic Vehicle Routing and Dispatching Jean-Yves Potvin Département d informatique et recherche opérationnelle and Centre interuniversitaire de recherche sur les réseaux d entreprise, la logistique et

More information

An Efficient Scheduling Algorithm for Multiple Charge Migration Tasks in Hybrid Electrical Energy Storage Systems

An Efficient Scheduling Algorithm for Multiple Charge Migration Tasks in Hybrid Electrical Energy Storage Systems An Efficient Scheduling Algorithm for Multiple Charge Migration Tasks in Hybrid Electrical Energy Storage Systems Qing Xie, Di Zhu, Yanzhi Wang, and Massoud Pedram Department of Electrical Engineering

More information

ENHANCED ANT COLONY ALGORITHM FOR GRID SCHEDULING

ENHANCED ANT COLONY ALGORITHM FOR GRID SCHEDULING ENHANCED ANT COLONY ALGORITHM FOR GRID SCHEDULING D.Maruthanayagam Head, Department of computer Applications, Vivekanandha College of Arts Sciences for Women, Tiruchengode, Tamilnadu, India (kesimaruthu@gmail.com)

More information

Operations and Supply Chain Management Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras

Operations and Supply Chain Management Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras Operations and Supply Chain Management Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras Lecture - 24 Sequencing and Scheduling - Assumptions, Objectives and Shop

More information

A Genetic Algorithm Applying Single Point Crossover and Uniform Mutation to Minimize Uncertainty in Production Cost

A Genetic Algorithm Applying Single Point Crossover and Uniform Mutation to Minimize Uncertainty in Production Cost World Applied Sciences Journal 23 (8): 1013-1017, 2013 ISSN 1818-4952 IDOSI Publications, 2013 DOI: 10.5829/idosi.wasj.2013.23.08.956 A Genetic Algorithm Applying Single Point Crossover and Uniform Mutation

More information

CHAPTER 4 PROPOSED HYBRID INTELLIGENT APPROCH FOR MULTIPROCESSOR SCHEDULING

CHAPTER 4 PROPOSED HYBRID INTELLIGENT APPROCH FOR MULTIPROCESSOR SCHEDULING 79 CHAPTER 4 PROPOSED HYBRID INTELLIGENT APPROCH FOR MULTIPROCESSOR SCHEDULING The present chapter proposes a hybrid intelligent approach (IPSO-AIS) using Improved Particle Swarm Optimization (IPSO) with

More information

Optimizing Grid-Based Workflow Execution

Optimizing Grid-Based Workflow Execution Journal of Grid Computing (2006) 3: 201 219 # Springer 2006 DOI: 10.1007/s10723-005-9011-7 Optimizing Grid-Based Workflow Execution Gurmeet Singh j, Carl Kesselman and Ewa Deelman Information Sciences

More information

Single machine scheduling with two agents for total completion time objectives

Single machine scheduling with two agents for total completion time objectives Lecture Notes in Management Science (2016) Vol. 8, 106 112 ISSN 2008-0050 (Print), ISSN 1927-0097 (Online) Single machine scheduling with two agents for total completion time objectives Yuvraj Gajpal 1

More information

LOAD SHUFFLING AND TRAVEL TIME ANALYSIS OF A MINILOAD AUTOMATED STORAGE AND RETRIEVAL SYSTEM WITH AN OPEN-RACK STRUCTURE

LOAD SHUFFLING AND TRAVEL TIME ANALYSIS OF A MINILOAD AUTOMATED STORAGE AND RETRIEVAL SYSTEM WITH AN OPEN-RACK STRUCTURE LOAD SHUFFLING AND TRAVEL TIME ANALYSIS OF A MINILOAD AUTOMATED STORAGE AND RETRIEVAL SYSTEM WITH AN OPEN-RACK STRUCTURE Mohammadreza Vasili *, Seyed Mahdi Homayouni * * Department of Industrial Engineering,

More information

An Optimal Job Scheduling in Grid Using Cuckoo Algorithm

An Optimal Job Scheduling in Grid Using Cuckoo Algorithm International Journal of Computer Science and Telecommunications [Volume 3, Issue 2, February 2012] 65 ISSN 2047-3338 An Optimal Job Scheduling in Grid Using Cuckoo Algorithm M. Prakash 1, R.Saranya 2,

More information

Queue Waiting Time Aware Dynamic Workflow Scheduling in Multicluster Environments

Queue Waiting Time Aware Dynamic Workflow Scheduling in Multicluster Environments Yu ZF, Shi WS. Queue waiting time aware dynamic workflow scheduling in multicluster environments. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 25(4): 864 873 July 2010. DOI 10.1007/s11390-010-1067-6 Queue

More information

A PROFIT MAXIMIZATION SCHEME WITH GUARANTEED QUALITY OF SERVICE IN CLOUD COMPUTING

A PROFIT MAXIMIZATION SCHEME WITH GUARANTEED QUALITY OF SERVICE IN CLOUD COMPUTING A PROFIT MAXIMIZATION SCHEME WITH GUARANTEED QUALITY OF SERVICE IN CLOUD COMPUTING S.Jaya Bharathi, PG Scholar, Dept Of Computer Science Engineering,Mailam Engineering College, Villupuram. Dr.T.Priyaradhikadevi,

More information

SCHEDULING AND CONTROLLING PRODUCTION ACTIVITIES

SCHEDULING AND CONTROLLING PRODUCTION ACTIVITIES SCHEDULING AND CONTROLLING PRODUCTION ACTIVITIES Al-Naimi Assistant Professor Industrial Engineering Branch Department of Production Engineering and Metallurgy University of Technology Baghdad - Iraq dr.mahmoudalnaimi@uotechnology.edu.iq

More information

Research Article Integrated Production-Distribution Scheduling Problem with Multiple Independent Manufacturers

Research Article Integrated Production-Distribution Scheduling Problem with Multiple Independent Manufacturers Mathematical Problems in Engineering Volume 2015, Article ID 579893, 5 pages http://dx.doi.org/10.1155/2015/579893 Research Article Integrated Production-Distribution Scheduling Problem with Multiple Independent

More information

Application-adaptive resource scheduling in a computational grid *

Application-adaptive resource scheduling in a computational grid * 1634 Luan et al. / J Zhejiang Univ SCIENCE A 26 7(1):1634-1641 Journal of Zhejiang University SCIENCE A ISSN 19-395 (Print); ISSN 1862-1775 (Online) www.zju.edu.cn/jzus; www.springerlink.com E-mail: jzus@zju.edu.cn

More information

Designing an Effective Scheduling Scheme Considering Multi-level BOM in Hybrid Job Shop

Designing an Effective Scheduling Scheme Considering Multi-level BOM in Hybrid Job Shop Proceedings of the 2012 International Conference on Industrial Engineering and Operations Management Istanbul, Turkey, July 3 6, 2012 Designing an Effective Scheduling Scheme Considering Multi-level BOM

More information

The Design of SMART: A Scheduler for Multimedia Applications

The Design of SMART: A Scheduler for Multimedia Applications The Design of SMART: A Scheduler for Multimedia Applications Jason Nieh and Monica S. Lam {nieh,lam}@cs.stanford.edu Technical Report CSL-TR-96-697 Computer Systems Laboratory Stanford University June

More information

Pricing with Bandwidth Guarantees for Clients with multi-isp Connections

Pricing with Bandwidth Guarantees for Clients with multi-isp Connections Pricing with Bandwidth Guarantees for Clients with multi-isp Connections Rohit Tripathi and Gautam Barua Department of Computer Science and Engineering Indian Institute of Technology, Guwahati Guwahati-78039,

More information

A Variable Capacity Parallel Machine Scheduling Problem

A Variable Capacity Parallel Machine Scheduling Problem Proceedings of the 2012 International Conference on Industrial Engineering and Operations Management Istanbul, Turkey, July 3 6, 2012 A Variable Capacity Parallel Machine Scheduling Problem Emine Akyol

More information

Generational and steady state genetic algorithms for generator maintenance scheduling problems

Generational and steady state genetic algorithms for generator maintenance scheduling problems Generational and steady state genetic algorithms for generator maintenance scheduling problems Item Type Conference paper Authors Dahal, Keshav P.; McDonald, J.R. Citation Dahal, K. P. and McDonald, J.

More information

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution

More information