• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.1
• /
• pp.287-307
• /
• 2018

There are two targets to schedule parallel jobs in the Cloud: (1) scheduling the jobs as many as possible, and (2) reducing the average execution time of the jobs. Most of previous work mainly focuses on the computing speed of resources without considering other attributes, such as bandwidth, memory and so on. Especially, past work does not consider the supply-demand condition from those attributes. Resources have different attributes, considering those attributes together makes the scheduling problem more difficult. This is the problem that we try to solve in this paper. First of all, we propose a new parallel job scheduling method based on a classification method of resources from different attributes, and then a scheduling method-CPLMT (Cloud parallel scheduling based on the lists of multiple attributes) is proposed for the parallel tasks. The classification method categories resources into different kinds according to the number of resources that satisfy the job from different attributes of the resource, such as the speed of the resource, memory and so on. Different kinds have different priorities in the scheduling. For the job that belongs to the same kinds, we propose CPLMT to schedule those jobs. Comparisons between our method, FIFO (First in first out), ASJS (Adaptive Scoring Job Scheduling), Fair and CMMS (Cloud-Minmin) are executed under different environments. The simulation results show that our proposed CPLMT not only reduces the number of unfinished jobs, but also reduces the average execution time.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2000.04a
• /
• pp.90-93
• /
• 2000

This paper presents job scheduling for non-identical parallel machines using Simulated Annealing (SA). The scheduling problem accounts for allotting work parts of L lots into M parallel machines, where each lot is composed of N homogeneous jobs. Some lots may have different jobs while every job within each lot has common due date. Each machine has its own performance and set up time according to the features of the machine, and also by job types. A meta-heuristic, SA, is applied in this study to determine the job sequences of the scheduling problem so as to minimize total tardiness of due. The SA method is compared with a conventional steepest descent(SD) algorithm that is a typical tool for finding local optimum. The comparison shows the SA is much better than the SD in terms tardiness while SA takes longer , but acceptable time.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.11a
• /
• pp.630-633
• /
• 2007

클러스터를 이용하여 다수의 작업을 실행시키는 경우에 효율적으로 사용자들이 자원을 사용하기 위해서는 작업 스케줄링이 매우 중요하다. 다양한 스케줄링 방법들이 제안되었으며 그 중 효율적으로 병렬 작업을 스케쥴링하기 위해 제안된 방법으로는 backfilling, co-scheduling, gang scheduling을 들 수 있다. 이러한 연구에서는 이론적인 논의가 많았고, 실제로 구현을 했다고 하더라고 multiprocessor 를 대상으로 backfilling 을 다룬 내용이 대부분이었다. 이 논문은 클러스터상에서의 parallel job scheduling 에 대해 조사하였다.

• IE interfaces
• /
• v.12 no.4
• /
• pp.617-624
• /
• 1999

In order to maximize an availability of machine and utilization of space, the parallel machines scheduling problem with space limit is frequently discussed in the industrial field. In this paper, we consider a scheduling problem for assembly machine in ship engine assembly shop. This paper considers the parallel machine scheduling problem in which n jobs having different release times, due dates and space limits are to be scheduled on m parallel machines. The objective function is to minimize the sum of earliness and tardiness. To solve this problem, a heuristic is developed. The proposed heuristic is divided into three modules hierarchically: job selection, machine selection and job sequencing, solution improvement. To illustrate its effectiveness, a proposed heuristic is evaluated with a large number of randomly generated test problems based on the field situation. Through the computational experiment, we determine the job selection rule that is suitable to the problem situation considered in this paper and show the effectiveness of our heuristic.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.27 no.2
• /
• pp.97-109
• /
• 2002

Manufacturing environments in the real world are subject to many sources of change and uncertainty, such as new job releases, job cancellations, a chance in the processing time or start time of some operation. Thus, the realistic scheduling method should Properly reflect these dynamic environment. Based on the release times of jobs, JSSP (Job Shoe Scheduling Problem) can be classified as static and dynamic scheduling problem. In this research, we mainly consider the dynamic JSSP with continually arriving jobs. The goal of this research is to develop an efficient scheduling method based on GA (Genetic Algorithm) to address dynamic JSSP. we designed scheduling method based on SGA (Sing1e Genetic Algorithm) and PGA (Parallel Genetic Algorithm) The scheduling method based on GA is extended to address dynamic JSSP. Then, This algorithms are tested for scheduling and rescheduling in dynamic JSSP. The results is compared with dispatching rule. In comparison to dispatching rule, the GA approach produces better scheduling performance.

• Journal of information and communication convergence engineering
• /
• v.9 no.4
• /
• pp.471-474
• /
• 2011

In parallel job scheduling, each job can be executed simultaneously on multiple machines at a time. Thus in the input instance, a job $J_i$ requires the number $m_i$ of machines on which it shall be processed. The algorithm should determine not only the execution order of jobs but also the machines on which the jobs are executed. In this paper, when the jobs have deadlines, the problem is to maximize the total work of jobs which is completed by their deadlines. The problem is known to be strongly NP-hard [5] and we investigate the approximation algorithms for the problem. We consider a model in which the algorithm can have more machines than the adversary. With this advantage, the problem is how good solution the algorithm can produce against the optimal algorithm.

• Management Science and Financial Engineering
• /
• v.11 no.2
• /
• pp.19-43
• /
• 2005

This paper considers a variation of the customer order scheduling problem, and the variation is the case where the machine-job assignment is fixed. We examine the parallel machine environment, and the objective is to minimize the sum of the completion times of the batches. While a machine can process only one job at a time, different machines can simultaneously process different jobs in a batch. The recognition version of this problem is known to be NP-complete in the strong sense even if there exist only two parallel machines. When there are an arbitrary number of parallel machines, we establish three lower bounds and develop a dynamic programming (DP) algorithm which runs in exponential time on the number of batches. We present two simple but intuitive heuristics, SB and GR, and find some special cases where SB and GR generate an optimal schedule. We also find worst case upper bounds on the relative error. For the case of the two parallel machines, we show that GR generates an optimal schedule when processing times of all batches are equal. Finally, the heuristics and the lower bounds are empirically evaluated.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.1
• /
• pp.39-60
• /
• 2017

In the Cloud, for the scheduling of parallel jobs, there are many tasks in a job and those tasks are executed concurrently on different VMs (Visual machines), where each task of the job will be executed synchronously. The goal of scheduling is to reduce the execution time and to keep the fairness between jobs to prevent some jobs from waiting more time than others. We propose a Cloud model which has multiple Clouds, and under this model, jobs are in different lists according to the waiting time of the jobs and every job has different parallelism. At the same time, a new method-ZOMT (the scheduling parallel tasks based on ZERO-ONE scheduling with multiple targets) is proposed to solve the problem of scheduling parallel jobs in the Cloud. Simulations of ZOMT, AFCFS (Adapted First Come First Served), LJFS (Largest Job First Served) and Fair are executed to test the performance of those methods. Metrics about the waiting time, and response time are used to test the performance of ZOMT. The simulation results have shown that ZOMT not only reduces waiting time and response time, but also provides fairness to jobs.

• Korean Management Science Review
• /
• v.20 no.1
• /
• pp.51-64
• /
• 2003

The JSSP (Job Shop Scheduling Problem) Is one of the most general and difficult of all traditional scheduling problems. The goal of this research is to develop an efficient scheduling method based on genetic algorithm to address JSSP. we design scheduling method based on SGA (Single Genetic Algorithm) and PGA (Parallel Genetic Algorithm). In the scheduling method, the representation, which encodes the job number, is made to be always feasible, initial population is generated through integrating representation and G&T algorithm, the new genetic operators and selection method are designed to better transmit the temporal relationships in the chromosome, and island model PGA are proposed. The scheduling method based on genetic algorithm are tested on five standard benchmark JSSPs. The results were compared with other proposed approaches. Compared to traditional genetic algorithm, the proposed approach yields significant improvement at a solution. The superior results indicate the successful Incorporation of generating method of initial population into the genetic operators.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.17 no.3
• /
• pp.1-12
• /
• 1992

Parallel Genetic Algorithms (GAs) are developed to solve a single machine n-job scheduling problem which is to minimize the sum of absolute deviations of completion times from a common due date. (0, 1) binary scheme is employed to represent the n-job schedule. Two selection methods, best individual selection and simple selection are examined. The effect of crossover operator, due date adjustment mutation and due date adjustment reordering are discussed. The performance of the parallel genetic algorithm is illustrated with some example problems.