SURVEY OF JOB GROUPING BASED SCHEDULING IN GRID COMPUTING

(1)

Available Online at www.ijpret.com 234

INTERNATIONAL JOURNAL OF PURE AND

APPLIED RESEARCH IN ENGINEERING AND

TECHNOLOGY

A PATH FOR HORIZING YOUR INNOVATIVE WORK

SURVEY OF JOB GROUPING BASED SCHEDULING IN GRID COMPUTING

JAYASHREE S. CHIRDE1_{, S. R. JADHAV}2

1. ME Student, Department of CSE, Babasaheb Naik College of Engg, Pusad (INDIA). 2. Asst. Prof., Department of CSE, Babasaheb Naik College of Engg, Pusad (INDIA).

Accepted Date: 05/03/2015; Published Date: 01/05/2015

\

Abstract:Grid computing is a form of distributed computing that provides a platform for executing large-scale resource intensive applications on a number of heterogeneous computing systems across multiple administrative domains. Therefore, Grid platforms enable sharing, exchange, discovery, selection, and aggregation of distributed heterogeneous resources such as computers, databases and visualization devices. Job and resource scheduling is one of the key research area in grid computing. Job scheduling is used to schedule the user jobs to appropriate resources in grid environment. The goal of scheduling is that it achieves highest possible system throughput and match the application need with the available computing resources. In this paper, we will review the definition of grid computing, types of Grids, architecture of Grid computing, characteristics of Computational Grid and job grouping. Grid is a system in which machines are distributed across various organizations. It involves sharing of resources that are heterogeneous and geographically distributed to solve various complex problems and develop large scale applications. Grid computing is broad in its domain of application and raises research questions that span many areas of distributed computing and of computer science in general. In this paper, we will explain job grouping and resource scheduling algorithms that will benefit the researchers to carry out their further work in this area of research.

Keywords: Grid computing, Job grouping, Job Scheduling

Corresponding Author: MS. JAYASHREE S. CHIRDE

Access Online On:

www.ijpret.com

How to Cite This Article:

Jayashree S. Chirde, IJPRET, 2015; Volume 3 (9): 234-242

(2)

Available Online at www.ijpret.com 235 INTRODUCTION

The emergence of high speed networks has made it possible to share geographically distributed resources such as supercomputers, storage systems, databases and scientific instruments in order to gather, process and transfer data smoothly across different administrative domains. Aggregations of such distributed resources, called computational grids provide computing power that has made it possible to solve large scale problems in science, engineering and commerce.

Grid Computing allow sharing of resources from heterogeneous and distributed locations. Grid Computing has wide variety of application areas including science, medical and research areas. But there are also some challenges which arise in the environment of grid computing [1].

Grid computing has emerged as a distributed methodology that coordinates the resources that are spread in the heterogeneous distributed environment [2].

GRID computing has become apparent as the next generation parallel and distributed computing methodology. Its instance is to provide a service-oriented infrastructure to enable easy access to and coordinated sharing of geographically distributed resources for solving various kinds of large-scale parallel applications. Now a days, grid computing has been widely accepted, study, and given attention to by researchers [3]. Unlike the traditional file exchange, as supported by the Web or peer-to-peer systems, users in the grid can access the required resource or service in a transparent way as if they were to use local resources or services. However, it gives rise to any of two or more ideas conflict between grid users and resource providers in usage policy of the local resources. For users, in addition to simplicity and easiness, to get desirable service functionalities, some quality of service (QoS) targets associated with the service, such as grid service reliability [4].

TYPES OF GRIDS

(3)

data grid project and Globus are data grid initiatives, working on developing large scale data organizations. A service grid provides services that cannot be accomplished by any single machine. The service grid further categorized into on-demand grid, collaborative grid and multimedia grid systems [5].

Fig. 1 Different categories of Grid systems

III.ARCHITECTURE OF GRID COMPUTING

Grids provide protocols and services at five different layers as identified in the Grid protocol architecture (see Fig2)At the Fabric layer, Grids provide access to different resource types such as compute, storage and network resource, code repository, etc. Grids usually rely on existing fabric components, for instance, local resource managers.

General-purpose components such as GARA (general architecture for advanced reservation) , and specialized resource management services such as Falkon Connectivity layer defines core communication and authentication protocols for easy and secure network transactions. The GSI (Grid Security Infrastructure) protocol underlies every Grid transaction.

The Resource layer defines protocols for the publication, discovery, negotiation, monitoring, accounting and payment of sharing operations on individual resources.

The GRAM (Grid Resource Access and Management) protocol is used for allocation of computational resources and for monitoring and control of computation on those resources, and GridFTP for data access and high-speed data transfer.

(4)

services, and MPICH for Grid enabled programming systems, and CAS (community authorization service) for global resource policies. Fig 2. Grid Protocol Architecture

The Application layer comprises whatever user applications built on top of the above protocols and APIs and operate in VO environments.

Fig2: A typical view of Grid architecture

IV. CHARACTERISTICS OF COMPUTATIONAL GRID

There are many desirable properties and features that are required by a grid to provide users with a computing environment. They are as follows:

• Heterogeneity:-The grid involves a number of resources that are varied in nature and can encompass a large geographical distance through various domains.

• Scalability:-The grid should be tolerant to handle a large number of nodes without any performance degradation.

• Adaptability or Fault Tolerant:-In a grid unexpected computational aborts, hardware or software faults etc are high. These faults are generally handled by Resource Managers.

(5)

V. JOB SCHEDULING

Generally the objective of job scheduling is to have good load balancing among the processors, whereas for the later minimization of overall execution time is the main concern [8]. Job scheduling and resource scheduling are the two main necessities in grid computing. In job scheduling, the job scheduler has to find the appropriate resource for the job that the user submits [9]. It has to find the best machine in grid to process the user job.

Grid has two main schedulers such as local schedulers and grid schedulers. The local schedulers work in local computational environment and hence it is reliably, fast connection, works in uniform environment and also takes full control of the homogeneous resources [10]. Grid Schedulers also called as meta-schedulers are the top level schedulers. They are responsible for orchestrating resources that are managed by different local schedulers [11]. Scheduling can also be classified into static and dynamic scheduling. In static scheduling, before execution the jobs are assigned to the suitable machines and those machines will continue executing those jobs without interruption. In dynamic scheduling, the rescheduling of jobs is allowed. The jobs executing can be migrated based on the dynamic information about the workload of the resources [12]. In grid, there may be lots of resources to run a job. The main focus is to find the appropriate resource for the job that is to schedule the job. The methods for job scheduling are centralized, hierarchical and decentralized.

• In centralized scheduling, there will be a centralized scheduler and it is responsible for scheduling the jobs. It is very useful when all the resources have same objective.

• In hierarchical, there will be central scheduler. All jobs will be submitted to the central scheduler. The central scheduler redirects the jobs to the global scheduler.

• In decentralized, there is no central scheduler. Distributed schedulers coordinate with each other to schedule jobs

VI.DIFFERENT METHODS OF JOB GROUPING ALGORITHMS.

This section describes many scheduling algorithms for computational grid tasks. These algorithms mainly developed to reduce makes pan and processing time. A job-grouping based scheduling which does not consider dynamic characteristics and utilization of the resources [Liu and Liao (2009)]. A greedy meta-scheduling algorithm based on multiple simultaneous requests. Scheduler identifies the sites that can start the job earliest.

(6)

[Subramani et al. (2002)]. Muthuvelu et al. (2005) presents dynamic Job grouping strategy which concentrates on maximizing the utilization of Grid resource processing capabilities and reducing the overhead time and cost taken to execute the jobs through a batch mode dynamic scheduling. But this algorithm did not consider the user demand of each task and did not improve the user satisfaction.

Selvarani and Sudha Sadhasivam (2010) developed a heuristic approach based on Particle Swarm Optimization algorithm for scheduling tasks in grid environment. This scheduling algorithm groups the tasks in nonuniform manner and scheduling is done based on the processing capability of the resources. By grouping the jobs, this approach optimizes computation/communication ratio and the utilization of resources is also increased. User demands of the tasks are not considered for scheduling the tasks.

Keat and Fong (2006) developed an algorithm in which grouping of independent jobs with small processing requirements into suitable jobs with large processing requirements and considered network bandwidth into account for scheduling, but it did not consider the dynamic characteristics of resources, sufficient utilization of the resources and user demand of the tasks.

Jie Lin et al. (2007) observed that an application demand aware performs better when user satisfaction is taken into account but data requirement and dynamic characteristics of the resources are not considered.

Suresh et al. (2011) developed a prioritized user demand algorithm which considered user deadline for allocating jobs from different users to different heterogeneous resources from different administrative domains.

In this algorithm better makespan and more user satisfaction is achieved but data requirement is not considered.

Soni et al. (2010) proposed a job scheduling algorithm to group the light-weight or small jobs into a coarse grained

or group of jobs, which will reduce the communication time, processing time and enhance resource utilization. But it is not concentrated on user satisfaction.

(7)

Applications. This algorithm considered computation scheduling and data scheduling are independent; they do not incorporate with each other to get suitable resources for a job.

Suresh and Balasubramanie (2012) presented user demand aware scheduling algorithm for data intensive tasks which considers user deadline, execution time and communication time for scheduling the tasks to resources. But dynamic characteristics of resources and processing overhead are not considered in this algorithm.

In Highest Response Next Scheduling schema [14] jobs are allotted to number of processors based on job’s priority and processor’s capability. HRN scheduling algorithm was proposed to correct the weaknesses of both Shortest Job First and FCFS. This algorithm provided for response with time, memory and CPU requirement. The advantages are it utilizes the resources efficiently and HRN model is much adaptive for

Grid environment. The disadvantages are that it contains high turnaround time, memory and CPU wastage.

ORC scheduling algorithm includes the combination of both the Best fit allocation and Round Robin scheduling to allocate the jobs in queue pool [15]. This algorithm improved the efficiency of load balancing and dynamicity capability of the grid resources. The advantages are reduces waiting time of job and turnaround time and increases processing time of jobs. The disadvantages are high communication overhead.

CONCLUSION

Grid computing can be solve complex tasks in less time and utilizes the hardware efficiently. The best job scheduling strategies have to be employed to make the grid work efficiently. Job group based scheduling is the foremost step in grid computing where the users’ jobs are group and scheduled to different machines. The various strategies have been studied and classified. In this paper we have discussed various job scheduling algorithms and job grouping algorithms that can be used to schedule different jobs. This paper also describes the various components of the grid scheduling system.

REFERENCES

1. Dipti Sharma ,Mr. Pradeep Mittal, 2013”Job Scheduling Algorithm for Computational Grid in

(8)

2. Mrs. Radha, Dr. V. Sumathy “A Detailed Study of Resource Scheduling and Fault Tolerance in

Grid”, 2011.

3. I. Foster, 2002, “The Grid: A new infrastructure for 21st century science,” Physics Today.

4. Y. S. Dai, M. Xie, and K. L. Poh, 2006, “Reliability of grid service systems,” Computers and

Industrial Engineering.

5. S. C. Guo, H. Wan, G. B. Wang, and M. Xie, 2010, “Analysis of grid resource compensation in

market-oriented environment,” Eksploatacja INiezawodnooeæ—Maintenance and Reliability.

6. Muthuvelu.N, Liu.J, Soe.N.L, Venugopal.S, Sulistio.A, and Buyya.R, (2005). A dynamic job

grouping-based scheduling for deploying applications with fine-grained tasks on global grids, in Proc of Australasian workshop on grid computing”,vol.4,pp.41–48.

7. http://www.ibrarian.net/navon/page.jsp?paperid=1548903&searchTerm=de+dialog”.

8. Ranganathan. K and Foster.I, (2002). Decoupling Computation and Data Scheduling in

Distributed Data-Intensive Applications,

9. Proceedings of 11th IEEE International Symposium on High Performance Distributed

Computing (HPDC-11), Edinburgh, Scotland, July 2002, IEEE CS Press,

10.Selvarani.S and Sudha sadhasivam.G, (2010). Improved job-grouping based pso algorithm

for task scheduling in grid computing,

11.International Journal of Engineering Science and Technology Vol.2(9),4687-4695.

www.ijest.info/docs/IJEST10-02-09-144.pdf

12.Soni.V.K,, Sharma.R and Mishra.M.K, (2010). Grouping-Based Job Scheduling Model In Grid

Computing, World Academy of

13.Science, Engineering and Technology, “http://www.waset.org/journals/waset/v41.php”

14.Subramani.V, Kettimuthu.R, Srinivasan.S, and Sadayappan.P, (2002). Distributed Job

Scheduling on Computational Grids Using

15.Multiple Simultaneous Requests, In Proceedings of the 11th IEEE Symposium on High

(9)

16.R. Buyya and M.Murshed, 2002, “Gridsim: a toolkit for the modeling and simulation of

distributed resource management and scheduling for grid computing,” Concurrency and Computation: Practice and Experience.

17.Mr. Rakesh Kumar and Navjot Kaur, “Job Scheduling in Grid Computers”.

18.Khushboo Yadav, Deepika Jindal, Ramandeep Singh 2013 “Job Scheduling in Grid

Computing” International Journal of Computer Applications (0975 – 8887) Volume 69– No.22, May 2013

19.K. Somasundaram, S. Radhakrishnan M. Gomathynayagam, 2007, “Efficient Utilization of

Computing Resources using Highest Response Next Scheduling in Grid” 6 (5): 544-547, Asian Journal of Information Technology.

20.K. Somasundaram, S.Radhakrishnan, 2008, “Node Allocation in Grid Computing Using

Optimal Resource Constraint (ORC) Scheduling”,VOL.8 No.6, IJCSNS International Journal of Computer Science and Network Security.

21.J. Santoso; G.D. van Albada; B.A.A. Nazief and P.M.A.Sloot, 2000, “Hierarchical Job

Scheduling for Clusters of Workstations”, ASCI 2000, pp. 99-105. ASCI, Delft.

22.Pinky Rosemarry, Ravinder Singh, Payal Singhal and Dilip Sisodia “GROUPING BASED JOB

SCHEDULING ALGORITHMUSING PRIORITY QUEUE AND HYBRID ALGORITHMIN GRID COMPUTING “IJGCA Vol.3, No.4, December 2012

23.P.SURESH, P.BALASUBRAMANIE,”GROUPING BASED USER DEMAND AWARE JOB