HOST SELECTION METHODOLOGY IN CLOUD COMPUTING ENVIRONMENT



Similar documents
HOST SCHEDULING ALGORITHM USING GENETIC ALGORITHM IN CLOUD COMPUTING ENVIRONMENT

A Data Placement Strategy in Scientific Cloud Workflows

Review of Cloud Computing Architecture for Social Computing

An Efficient Checkpointing Scheme Using Price History of Spot Instances in Cloud Computing Environment

Cost Efficient Datacenter Selection for Cloud Services

Firewall Design: Consistency, Completeness, and Compactness

Agent Based Framework for Scalability in Cloud Computing

Round Robin with Server Affinity: A VM Load Balancing Algorithm for Cloud Based Infrastructure

International Journal of Computer Science Trends and Technology (IJCST) Volume 2 Issue 3, May-Jun 2014

State of Louisiana Office of Information Technology. Change Management Plan

Minimizing Makespan in Flow Shop Scheduling Using a Network Approach

Performance Analysis of VM Scheduling Algorithm of CloudSim in Cloud Computing

High performance computing network for cloud environment using simulators

On Adaboost and Optimal Betting Strategies

GPRS performance estimation in GSM circuit switched services and GPRS shared resource systems *

How To Segmentate An Insurance Customer In An Insurance Business

CloudAnalyzer: A cloud based deployment framework for Service broker and VM load balancing policies

Bellini: Ferrying Application Traffic Flows through Geo-distributed Datacenters in the Cloud

Unbalanced Power Flow Analysis in a Micro Grid

Heterogeneous Workload Consolidation for Efficient Management of Data Centers in Cloud Computing

Modelling and Resolving Software Dependencies

A Review of Load Balancing Algorithms for Cloud Computing

DECISION SUPPORT SYSTEM FOR MANAGING EDUCATIONAL CAPACITY UTILIZATION IN UNIVERSITIES

Beyond the Internet? THIN APPS STORE FOR SMART PHONES BASED ON PRIVATE CLOUD INFRASTRUCTURE. Innovations for future networks and services

Load Balancing using DWARR Algorithm in Cloud Computing

Improving Emulation Throughput for Multi-Project SoC Designs

Forecasting and Staffing Call Centers with Multiple Interdependent Uncertain Arrival Streams

ThroughputScheduler: Learning to Schedule on Heterogeneous Hadoop Clusters

Analysis of Service Broker Policies in Cloud Analyst Framework

Towards a Framework for Enterprise Architecture Frameworks Comparison and Selection

SLA-aware Resource Scheduling for Cloud Storage

A New Evaluation Measure for Information Retrieval Systems

Enterprise Resource Planning

Efficient Qos Based Resource Scheduling Using PAPRIKA Method for Cloud Computing

Detecting Possibly Fraudulent or Error-Prone Survey Data Using Benford s Law

10.2 Systems of Linear Equations: Matrices

Stock Market Value Prediction Using Neural Networks

Presentation of Multi Level Data Replication Distributed Decision Making Strategy for High Priority Tasks in Real Time Data Grids

USING SIMPLIFIED DISCRETE-EVENT SIMULATION MODELS FOR HEALTH CARE APPLICATIONS

Geoprocessing in Hybrid Clouds

Energy Cost Optimization for Geographically Distributed Heterogeneous Data Centers

Optimal Energy Commitments with Storage and Intermittent Supply

Simulation of Boiler Model in a Cloud Environment

Dynamic Network Security Deployment Under Partial Information

An intertemporal model of the real exchange rate, stock market, and international debt dynamics: policy simulations

The one-year non-life insurance risk

AN IMPLEMENTATION OF E- LEARNING SYSTEM IN PRIVATE CLOUD

A Service Revenue-oriented Task Scheduling Model of Cloud Computing

Resource Allocation Avoiding SLA Violations in Cloud Framework for SaaS

An Efficient Approach for Task Scheduling Based on Multi-Objective Genetic Algorithm in Cloud Computing Environment

Unified API Governance in the New API Economy

Minimum-Energy Broadcast in All-Wireless Networks: NP-Completeness and Distribution Issues

Product Differentiation for Software-as-a-Service Providers

Data Center Power System Reliability Beyond the 9 s: A Practical Approach

Dynamic Resource Pricing on Federated Clouds

PRIVACY PRESERVATION ALGORITHM USING EFFECTIVE DATA LOOKUP ORGANIZATION FOR STORAGE CLOUDS

A Blame-Based Approach to Generating Proposals for Handling Inconsistency in Software Requirements

A NEW APPROACH FOR LOAD BALANCING IN CLOUD COMPUTING

MODIFIED BITTORRENT PROTOCOL AND ITS APPLICATION IN CLOUD COMPUTING ENVIRONMENT

JON HOLTAN. if P&C Insurance Ltd., Oslo, Norway ABSTRACT

! # % & ( ) +,,),. / % ( 345 6, & & & &&3 6

CLOUD COMPUTING: A NEW VISION OF THE DISTRIBUTED SYSTEM

A Survey on Load Balancing and Scheduling in Cloud Computing

Supply Chain Platform as a Service: a Cloud Perspective on Business Collaboration

SERVICE BROKER ROUTING POLICES IN CLOUD ENVIRONMENT: A SURVEY

Supporting Adaptive Workflows in Advanced Application Environments

Keywords: Cloudsim, MIPS, Gridlet, Virtual machine, Data center, Simulation, SaaS, PaaS, IaaS, VM. Introduction

RUNESTONE, an International Student Collaboration Project

Advanced Task Scheduling for Cloud Service Provider Using Genetic Algorithm

THE CLOUD AND ITS EFFECTS ON WEB DEVELOPMENT

An Experimental Study of Load Balancing of OpenNebula Open-Source Cloud Computing Platform

Utilizing Round Robin Concept for Load Balancing Algorithm at Virtual Machine Level in Cloud Environment

An introduction to the Red Cross Red Crescent s Learning platform and how to adopt it

Ch 10. Arithmetic Average Options and Asian Opitons

Data Integrity Check using Hash Functions in Cloud environment

Professional Level Options Module, Paper P4(SGP)

Seeing the Unseen: Revealing Mobile Malware Hidden Communications via Energy Consumption and Artificial Intelligence

International Journal of Computer Science Trends and Technology (IJCST) Volume 3 Issue 3, May-June 2015

Efficient and Enhanced Algorithm in Cloud Computing

CloudSim-A Survey on VM Management Techniques

Cross-Over Analysis Using T-Tests

A Comparative Study on Load Balancing Algorithms with Different Service Broker Policies in Cloud Computing

EFFICIENT VM LOAD BALANCING ALGORITHM FOR A CLOUD COMPUTING ENVIRONMENT

An Implementation of Load Balancing Policy for Virtual Machines Associated With a Data Center

Advanced Load Balancing Mechanism on Mixed Batch and Transactional Workloads

Consumer Referrals. Maria Arbatskaya and Hideo Konishi. October 28, 2014

A hybrid approach to supply chain modeling and optimization

Dow Jones Sustainability Group Index: A Global Benchmark for Corporate Sustainability

FAST JOINING AND REPAIRING OF SANDWICH MATERIALS WITH DETACHABLE MECHANICAL CONNECTION TECHNOLOGY

A Theory of Exchange Rates and the Term Structure of Interest Rates

View Synthesis by Image Mapping and Interpolation

Transcription:

International Journal of Avance Research in Computer Engineering & Technology (IJARCET) HOST SELECTION METHODOLOGY IN CLOUD COMPUTING ENVIRONMENT Pawan Kumar, Pijush Kanti Dutta Pramanik Computer Science & Engineering Department B. T. Kumaon Institute of Technology Dwarahat, Almora, (Uttarakhan), Inia Abstract Clou computing is a paraigm in which IT (information technology) application provie as a service. It allows users to utilize on-eman computation over internet, which is helpful for storage of ata an services from aroun the worl in commercialize manner. In clou environment, applications nee access to mass atasets that may each be replicate on ifferent resources (or ata centers) an Mass ata moving from user to host an hosts to user. Base on the above two points, how to select best host for accessing resources an creating a virtual machine(vm) to execute applications to make execution efficiency high an access cost low as far as possible simultaneously is a challenging an urgent problem. In this paper, a host selection moel base on minimum network elay using WSCP combine with Max-Min Heuristic. To select the host an scheule multiple jobs on multiple machines in an efficient manner is propose, the objective is to minimize propagation time of input an output ata by selecting nearest host into the network an finally it minimize the execution time of cloulet. Inex Terms clou Computing, WSCP, Max-Min Heuristic. I. INTRODUCTION Following istribute computing, parallel computing, gri computing, utility computing, Web 2.0. etc., the computer inustry an acaemia put forwar clou computing moel [1], which achieves generalization an commercialization of these previous moels in some sense [2]. Clou computing, the long-hel Manuscript receive Oct, 2012. Pawan Kumar, Computer Science an Engg., UTU/Bipin Chanra Tripathi Kumaon Institute of Technology. Almora, Inia. Pijush Kanti Dutta Pramanik, Computer Science an Engg., UTU/Bipin Chanra Tripathi Kumaon Institute of Technology. Almora, Inia. ream of computing, has the potential to change a large part of the IT inustry, making it even more attractive as a service an shaping the way IT harware is esigne an purchase[3]. No oubts it woul increasingly change the way people live an work. Clou computing can be efine as a type of parallel an istribute system consisting of a collection of interconnecte an virtualize computers that are ynamically provisione an presente as one or more unifie computing resources base on service-level agreements establishe through negotiation between the service proviers an consumers [1]. The clou computing is still at its infant stage an a very new technology for enterprises. Clou computing is term use to escribe both a platform an types of application. As a platform its supplies, configure an reconfigures servers, while servers can be physical machine or virtual machine. On the other han, clou computing escribes application that are extene to be accessible through the internet an for this large ata centers an powerful servers are use to host the web application an web services. There are some important points in the efinition to be iscusse regaring clou computing. Clou computing iffers from traition computing paraigm as it is scalable, can been capsulate as an abstract entity which provies ifferent level of services to the clients, riven by economies of scale an the service are ynamically configurable. As a very new technology for enterprises there are many benefits state of clou computing by ifferent researchers which make it more preferable to be aopte by enterprises. Clou computing infrastructure allows achieving more efficient use of there IT harware an software investments. This is achieve by breaking own the physical barrier inherent in isolate systems, automating the management of the group of the systems as a entity. All Rights Reserve 2012 IJARCET 1

International Journal of Avance Research in Computer Engineering & Technology (IJARCET) Clou computing can also be escribe as the ultimately virtualize system an a natural evolution for ata centers which offer automate systems management. The paper concentrates on the selection of resources, that is, to select a ata center for creating VM to submit the task an several other ata centers for accessing replicas require by the task. The metho for the problem aopte here is: firstly, to fin a set of ata centers for the task to access all the replicas require, an then to fin an appropriate ata center among them for creating a VM to execute the task. Here we select the one who has the minimum transfer time from other ata centers in the set of all. Our aim is to reuce ata transfer times an access cost of ata by selecting an appropriate set of ata centers. In this paper, we propose a host selection methoology using WSCP couple with Max-Min Heuristic for fining a set of ata centers such that every ata center selecte contains replicas require as more as possible for reucing transfer times, while the total access cost of these replicas is as low as possible. That is to select ata resources with lowest average access cost of replicas. The rest of the paper is organize as follows: Section II introuces previous work in replica selection strategy in ata-intensive environment. Section III introuces the Heuristic base Scheuling. Section IV introuces the evaluation moel use in this research. Section V having the WSCP Base Max- Min heuristic etails. Section VI having the simulation an result etails. Finally, the paper was conclue in section VII. II. RELATED WORK In this section, we present some backgroun knowlege an literature review on host selection metho. The problem of resource selection in istribute environment has receive lots of attention in recent years. In many previous works, resource selection refers to the selection of computational resource in gri environment. In [], the authors presente a resource selection moel using ecision theory for selecting the best machine to run a task. This paper presents a resource selection moel using ecision theory for combining these influential factors in the resource selection process. This moel eploying istribute an parallel processing for job execution preiction. They have presente appropriate functional behaviors an positive performance results. In [5], they propose an algorithm for resource selection problem of computational gris, base on the resource-availability preiction using frequent workloa patterns. Resource selection is an important issue of gri computing. However, most of the propose methos are not effective enough to resolve the problem of resource selection in gris. The reason behin is that these methos usually make use of current workloa state or short-term preiction in available CPU time to be the basis of resource selection while most of gri jobs require a long execution time. Recently, with the rapi evelopment of ata intensive computing, many researchers turne their attention to resource selection of ata-intensive environment, such as ata gri [6]. In ata-intensive environment, besies computational resources, resources to be selecte inclue ata resources selection, which is equivalent to replica selection in ata gri. In [8], the author propose Economy-Base File Replication Strategy for a Data Gri. It use an auction protocol to select the cheapest replica of a ata set by a job running on computing element, which is lack consieration of the selection of computational resource. In [9], the author propose the atacenter selection base on number of PE available in to the host. So that it selects that host which has maximum no of free PE. It oesn t consier elay of transferring ata. In this paper, Weighte Set Covering Problem (WSCP) base on Max-Min heuristic is propose. For the moel, author applies a WSCP base on Max- Min to prouce an approximately optimal resource set for each task. The result shows that WSCP base on Max-Min heuristic can prouce an approximately optimal solution in most cases to meet both execution efficiency an economic emans simultaneously, compare to other strategies largescale. III. HEURISTIC BASED SCHEDULING There are various scheuling techniques, but here we iscuss only two of them. A. MIN-MIN Min-Min begins with the set MT (MetaTask) of all unassigne tasks an has two stages. In the first stage, the set of minimum expecte completion time for each task in MT is foun. In the secon stage, the All Rights Reserve 2012 IJARCET 2

International Journal of Avance Research in Computer Engineering & Technology (IJARCET) task with the overall minimum expecte completion time from MT is chosen an assigne to the corresponing machine. Then this task is remove from MT an the process is repeate until all tasks in the MT are mappe. However, the Min- Min Algorithm first finishes the shorter tasks an then executes the long task. B. MAX-MIN Max-Min is almost same as the min-min algorithm except the following: in this after fining out the completion time, the minimum execution times are foun out for each an every task. Then among these minimum times the maximum value is selecte which is the maximum time among all the tasks on any resources. Then that task or jobs is scheule on the resource on which it takes the minimum time an the available time of that resource is upate for all the other tasks. The upating is one in the same manner as for the Min-Min. All the tasks are assigne resources by this proceure. IV. EVALUATION MODEL A clou computing environment can be consiere as a set of P ata centers D ={ 1, 2,..., M }, which are connecte by high spee Internet. For an application mae up of a set of N inepenent tasks or jobs J ={j 1,j 2,.,j N },(N>>M),each job j J, require a set of k ata set, enote by F j, that are sprea on a subset of D. Consier a set of N inepenent tasks(jobs) submitte to a VM, which is create on ata center D. This is shown below fig. no. 1 V. WSCP BASED MAX-MIN HEURISTIC In clou computing environment clou application can be consiere as a set of inepenent tasks (jobs), each of which require for job j to be submitte to a VM that is create on ata center D, the K atasets require, enote by F j, are sprea on m ifferent ata centers at ifferent costs, It can be translate into a form of ajacency matrix A = [a ik ], 1 i P, 1 k K wherein a ik = w ik (w ik > 0) if VM can access ataset f k from ata center i at a cost of w ik, which is abstracte as weight of f k ; an otherwise a ik = 0, that is i oesn t contain f k. The rows that contain a w ik in a particular column are sai to cover the column at cost of w ik. The problem of fining an optimal ata centers set such that each can cover replicas as more as possible, an the total access cost of replicas is as cheap as possible, can be consiere as the problem of fining a set of ata center, each of which has lowest average access cost of replicas. This problem is equivalent to the problem of fining an optimal set of rows to cover all the columns with the lowest average weight representing access cost. While the mapping heuristic fins a resource set for a single job, the overall objective is to minimize the total makespan, the total time from the start of the scheuling to the completion of the last job, of the application consisting of N such ataintensive jobs. At the en, we apply the wellknown Max-Min, propose by Maheswaran et al. [11], for ynamic scheuling of jobs on heterogeneous computing resources. Our whole algorithm is shown below. JOB SET DATA SET ALGORITHM1. WSCP BASED HEURISTIC V M j f 1 f 2 Fig1. Resource Selection Moel f k 1 1 2 a a s i D t 1 2 Begin Main All Rights 2 Reserve 2012 IJARCET 1. For a task j, create the ajacency matrix A with ata centers forming the rows an atasets forming the columns 2. Initial solution set B, E,L an z ; a ata center NULL 3. Search(L, T, B, E, z)). S j {{r}, L} where r R such that S j prouces MCT (B) En Main Search (L, T, B, E, z) 5. Fin the minimum k, such that f k E. Let T k be the block of rows in T corresponing to f k.. set a pointer q to the of T k 6. While q oesn t reach the en of T k o 3

International Journal of Avance Research in Computer Engineering & Technology (IJARCET) 7. F T { f i t qi =1,1 i k} 8. B BU { k q }, E EU F T 9. if E=F j 10. if z > MCT(B) then 11. L B, z MCT(B) 12. Else Search (L,T, B, E, z) 13. B B-{ k q }, E E- F T 1. Increment q En MCT (B) 15. Fin r R such that the completion time is minimum for the resource set S j ={{r},r} an return value efforts In ClouSim, users is moele by a DatacenterBroker, which is responsible for meiating between users an service proviers epening on user s tasks across Clous. In our experiments, we have use ClouSim as a simulator for checking the performance of our improve algorithm. We have consiere Virtual Machines as resource an Cloulets as tasks/jobs. We have checke the performance of the algorithm by fixe the number of virtual machines an varie the number of cloulets. The makespans that the algorithms prouce are shown in fig no. 2 we have fixe the number of virtual machines as 20 an we are varying the number of cloulets from 20 to 120 with a ifference of 20. ALGORITHM2. WSCP BASED MAX-MIN HEURISTIC Begin Main 1. Repeat 2. foreach j J u o 3. Fin the resource set by WSCP that achieve the MCT for j. en 5. Fin the job j J u with maximum value of T ct (j) 6. Assign j to its selecte resource set an remove j from j u 7. Upate the resource availability base on the allocation performe in the previous step 8. Until j u is empty 180 160 10 120 100 80 60 0 20 0 20 0 60 80 100120 scp wscp with maxmin En Main VI. SIMULATION AND RESULT ClouSim leae by Buyya, allows clou customers to test their services in repeatable an controllable environment free of cost, an to turn the performance bottlenecks before eploying on real clous. It can provie a generalize an extensible simulation framework that enables moeling, simulation an experimentation of emerging clou computing infrastructures an application services. It is esigne for stuying various resource management approaches an scheuling algorithms in clou environment. The ClouSim toolkit supports both system an behavior moeling of Clou system components such as ata centers, virtual machines (VMs) an provisioning policies of resource. It implements generic application provisioning techniques that can be extene with ease an limite Fig2. Graph for Makespans VII. CONCLUSION We have esigne an teste an algorithm which is mae by WSCP couple with Max-Min Heuristic. The main goal of it, to select the host an to scheule multiple jobs on multiple machines in an efficient manner such that the jobs take the minimum time for the completion. VIII. REFERENCES [1] I. Foster, Y Zhao, I. Raicu, an S. Lu, Clou Computing an Gri Computing 360- egreecompare[c], in Gri Computing Environments Workshop, 2008, pp. 1-10. [2] Daniel Nurmi, Rich Wolski, Chris Grzegorczyk, Graziano Obertelli, Sunil Soman,Lamia Youseff, Dmitrii Zagoronov, The Eucalyptus Open-source Cloucomputing System, 2009 9th IEEE/ACM International Symposium on Cluster Computing an the Gri, CCGRID 2009, pp: 12-131. All Rights Reserve 2012 IJARCET

International Journal of Avance Research in Computer Engineering & Technology (IJARCET) [3] Michael Armbrust, Armano Fox, Rean Griffith, Anthony D. Joseph, Rany H. Katz, Anrew Konwinski, Gunho Lee, Davi A. Patterson, Ariel Rabkin, Ion Stoica, Matei Zaharia, Above the Clous: A Berkeley View of Clou Computing, Technical Report No. UCB/EECS-2009-28, 2009. [] Lilian Noronha Nassif, José Marcos Nogueira, Flávio Vinícius e Anrae, Distribute Resource Selection in Gri Using Decision Theory, in Seventh IEEE International Symposium on Cluster Computing an the Gri(CCGri 2007,pp:97-102). [5] Tyng-Yeu Liang Siou-Ying Wang I-Han Wu, Using Frequent Workloa Patterns in Resource Selection for Gri Jobs, DOI 10.1109/APSCC.2008.217. [6] D.G. Cameron, R. Carvajal-Schiaffino, A.P. Millar, C. Nicholson, K.Stockinger, F. Zini, Evaluating scheuling an replica optimisation strategies in OptorSim, in:proceeings of the Fourth International Workshop on Gri Computing (Gri2003), IEEE CS Press, Los Alamitos,CA, USA, Phoenix, AZ, USA, 2003. [7] Rajkumar Buyya, Rajiv Ranjan, Rorigo N. Calheiros, Moeling an Simulation of Scalable Clou Computing Environments an the ClouSim Toolkit: Challenges an Opportunities, in The 2009 International Conference on High Performance Computing an Simulation, HPCS 2009, pp:1-11. [8] Juefu Liu, Peng Liu, Status an Key Techniques in Clou Computing, in Proceeings of 2010 3r International Conference on Avance Computer Theory an Engineering (ICACTE), pp: V-285 V- 288. [9] H. Baghban an M. Rahmani, A Heuristic on Job Scheuling in Gri Computing Environment, in Proc. 7th Inter. Conf. on Gri an Cooperative Computing (GCC 08), 2008, pp. 11-16. [10] Huang Q.Y., Huang T.L., An Optimistic Job Scheuling Strategy base on QoS for Clou Computing, IEEE International Conference on Intelligent Computing an Integrate Systems (ICISS), 2010, Guilin, pp. 673-675, 2010. [11] M. Maheswaran, S. Ali, H. J. Siegel, D. Hensgen an R. F. Freun, Dynamic Matching an Scheuling of a Class of Inepenent tasks onto Heterogeneous Computing Systems, Journal of Parallel an Distribute Computing, Vol. 59, No. 2, pp. 107-131,1999. [12] E. Munir, J. Li, S. Shi an Q. Rasool, Performance Analysis of Task Scheuling Heuristics in Gri, in Proc. 6th Inter. Conf. on Machine Learning an Cybernetics, 2007, pp. 3093-3098. All Rights Reserve 2012 IJARCET 5