An Improved Load Balancing Approach to Minimize Response Time using Genetic Algorithm in Cloud...

33
An Improved Load Balancing Approach to Minimize Response Time using Genetic Algorithm in Cloud Computing A Thesis Presentation Presented in partial fulfillment of the requirement for the award of degree of MASTER OF TECHNOLOGY IN COMPUTER SCIENCE & ENGINEERING Presented By MONIKA LAGWAL (0901CS15MT07) Under the supervision of PROF. NEHA BHARDWAJ Assistant Professor 2015-2017 DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING AND INFORMATION TECHNOLOGY MADHAV INSTITUTE OF TECHNOLOGY & SCIENCE, GWALIOR (M.P.) - 474005

Transcript of An Improved Load Balancing Approach to Minimize Response Time using Genetic Algorithm in Cloud...

An Improved Load Balancing Approach to Minimize Response Time using Genetic Algorithm in Cloud Computing

AThesis Presentation

Presented in partial fulfillment of the requirement for the award of degree ofMASTER OF TECHNOLOGY

INCOMPUTER SCIENCE & ENGINEERING

Presented ByMONIKA LAGWAL(0901CS15MT07)

Under the supervision ofPROF. NEHA BHARDWAJ

Assistant Professor

2015-2017DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING AND INFORMATION

TECHNOLOGYMADHAV INSTITUTE OF TECHNOLOGY & SCIENCE,

GWALIOR (M.P.) - 474005

Contents1. INTRODUCTION

1.1. CLOUD COMPUTING

1.2. TYPE OF CLOUD SERVICE MODELS

2. LOAD BALANCING

3. LITERATURE SURVEY

4. EXISTING METHODOLGY

5. PROBLEM STATEMENT

6. PROPOSED WORK

7. RESULTS

8. CONCLUSION AND FUTURE

9. PUBLICATIONS

10. REFERENCES

Cloud Computing

Cloud computing is a model for providing on-demand resources, network access to a

shared pool of configurable computing resources (e.g., networks, servers, storage,

applications, and services) that can be rapidly provisioned and released with minimal

management effort or service provider interaction.

Deployment Models of Cloud Computing

1. Public Clouds -e.g. Drop box, Gmail.

2. Private Clouds -e.g. IBM ,Window Azure and

3. Hybrid Clouds -e.g. Google Compute Engine.

TYPE OF CLOUD SERVICE MODELS

1.Software as a Service (e.g. Salesforce.com)

2.Platform as a Service (e.g. App Engine) and

3.Infrastructure as a Service (e.g. Amazon EC2 )

LOAD BALANCING

• Load balancing is technique to divide the amount of workload between two or more computers so that more work gets done in the same amount of time and, in general, all users get serve faster.

• Load balancing technique ensures that no node is idle while other nodes are being utilized.

• Load balancing can be implemented with hardware, software, or a combination of both.

CLOUD HOSTING ENVIRONMENT IN CLOUD LOAD BALANCING

Goals of Load-Balancing

• Achieve optimal resource utilization

• Maximize throughput

• Minimize response time

• Avoid overload

Literature survey

[1] Sakshi et al.[2016] The survey of different load balancing algorithms and their comparison with each other is done.

[2] Jayashree et al.[2015] The mainly focuses on various dynamic load balancing methods used for optimized resource management in cloud.

Literature survey[4] Ronak R Patel et al. [2016] Proposed a load-balancing mechanism based on improved GA using PR (Population Reduction Method) gives satisfactory results. After selecting resources by PR Method overload resource and transfer that load to ideal are handle. In paper, cloud-sim simulator for checking the allocation and load on resources. When are entered into cloud simulator then first population is define on population PRM method and after completing all jobs the response-time for resource will given an idea for finishing job time.

[5] Reena Panwar et al. [2015] A brief discussion on the existing load balancing techniques in cloud computing and further compares them based on various parameters like data processing time and response time etc. In paper cloud-analyzes toolkit used to obtain the result based on existing Round Robin and Throttled scheduling algorithms. Throttled LB technique is more efficient in terms of cost for LB on cloud data centers.

[6] Medhat Tawfeek et al. [2015] In paper, a cloud task-scheduling policy based on Ant-Colony Optimization algorithm which compared with dissimilar scheduling algorithms like First Come First Served (FCFS) and Round-Robin (RR), has been presented. The foremost goal of these algorithms is minimizing the make span of a given tasks set. ACO is haphazard optimization search-approach that will be used for assigning the incoming jobs to the virtual-machines. Algorithms have been simulated using cloud-sim toolkit package. Experimental results showed that cloud task scheduling based on ACO outperformed FCFS and R-R algorithms.

[7] Shang-Liang Chen et al. [2016] In paper, a load balancing algorithm CLB was proposed to balance the load among virtual machines in a cloud data center. The results showed that this algorithm can achieve better load balancing in a large scale cloud computing. .

LITERATURE SURVEY

AUTHOR YEAR PROPOSED WORK

PROBLEM STATEMENT

TECHNIQUE /TOOLKIT

RESULTS

[3] Shikha Garg et al.

2016 Enhanced-Active Monitoring Load-Balancing (EAMLB) algorithm

Response-time is not better.

Cloud-Analyst Min the Response-Time are gain to execute the jobs in queue

[4] Ronak R Patel et al.

2016 Improved GA using PR (Population Reduction Method)

Identifies and completes the job waiting in queue

Cloud-Sim Overload Resource and Transfer that load to ideal CPU

[5] Reena Panwar et al.

2015 Throttled LB technique

Provides better data processing -time, response time and cost

Cloud-Analyzes

Throttled LB technique is more efficient in terms of cost for LB on cloud data centers.

AUTHOR YEAR PROPOSED WORK

PROBLEM STATEMENT

TECHNIQUE /TOOLKIT

RESULTS

[6] Medhat Tawfeek et al.

2015 Ant-Colony Optimization algorithm

To minimizing the make span of a given tasks.

Cloud-Sim More efficient better results then exiting work.

[7] Shang-Liang Chen et al.

2016 Cloud Load Balancing algorithm

Response time and Migration- time is not done properly.

Cloud Load Balancing

Throughput, Overhead, fault tolerance, resource utilization and performance are handle in effective manner .

[14] Mr. M. Ajit et al.

2013 A new Weighted Signature based Load Balancing (WSLB)

VMs mapped on hosts in datacenter is not done effective way .

cloud-analyst Algorithm finds the load assignment factor for each of the host in a datacenter and maps the VMs according to that factor in effective way.

Existing Methodology In exiting work [7] they used cloud load balancing approach .

Step 1. With the help of weighted scheduling, tasks are separated (w1,w2,w3,w4) then

distributed to each VMs.

Step 2. In first round, VM1 fetch the first (tasks) process and start execution, then VM2

fetch the second (tasks) and this will cont. so on…

Step 3. In second round, if VM1 found too heavy loaded for the executed of process ,

then it will started , second round from VM2 and carry on the

execution.

In existing work the throughput , overhead , fault tolerance , resource utilization and performance are handle in effective manner but Migration and Response time is not done properly.

BASE PAPERThe cloud load balancing algorithm process.

PROBLEM STATEMENT

Response time and time efficient is not done properly.

Least weight jobs consider

Not time efficient

Not cost efficient

Here, every compile-time, the execution of cloudlets the result charging. Every-time we got different outputs result.

PROPOSED WORK

A new Genetic-algorithm approach is used with Cloud-sim tool, to over come from exiting problems, so that efficient Response-time can be gain.

A comparator analysis has been carried out with existing method; it is observed that proposed method give 28.53% improved response-time.

ALGORITHM

Input: Cloudlets, Virtual Machines.

Output: Effective Response-Time

Step 1. Start

Step 2. Select cloudlet(s)

Step 3. Short the cloudlets according to cost

// broker short the cloudlets on the basis of cost service

Step 4. If ((Distance&&Resource requirement&& Length) <=Threshold value)

// Selected threshold value [15]

// Distance between user and vm (service provider)

// Length i.e. number of instructions in the cloudlets

Cost effective cloudlets for VM

Else

Costly process // search next vm

Step 5. Now select shortlisted VM and Cloudlets // by fitness value

Step 6. Create chromosomes apply crossover operation

Step 7. If (vm have sufficient-space for cloudlets execution)

Cloudlets assign to VM

Else

Search next VM for cloudlets

Step 8. Repeat step from 2 to 8, all new arrival cloudlets

Step 9. Exit

Proposed Methodology

RESULTS OF BASE CODE WORKEXAMPLE 1:VM = 4CLOULDLET= 7 RESULTS=18.096 (milliseconds)

RESULTS OF PROPOSE CODE WORKEXAMPLE 1: VM = 4CLOULDLET= 7 RESULTS=7.599 ( ms)

RESULTS OF BASE CODE WORK EXAMPLE 2:VM = 8CLOULDLET= 18 RESULTS=16.0981 (milliseconds)

RESULTS OF PROPOSE CODE WORK EXAMPLE 2:VM = 8CLOULDLET= 18 RESULTS=14.0955 ( milliseconds)

Comparison Result between Existing and Proposed Techniques

PARAMETER

S.NO.

VM CLOUDLET(S) RESPONSE-TIME(in milliseconds)USINGCLB

RESPONSE -TIME (in milliseconds)USINGGA

1 4 7 18.096 7.599

2 8 18 16.07 14.08

3 6 13 15.09 8.22

Comparison of Base and Proposed Technique

Res

pons

e-tim

e in

(m

s)

CONCLUSION

With the help of a new design Genetic-algorithm approach is used with Cloud-sim tool, to over come from Exiting Problems (the Execution of Cloudlets in Minimum Response-Time).

In the simulation results show that 28.53% effective response time is gain with the help of 25 different simulations test have been tested.

FUTURE WORK

In future work, improve the policy and expand the algorithm to handle the Migration-Time.

Workflow-sim tool can be used to gain more better results.

LIST OF PUBLICATIONS(Presented/Published/Accepted)

1. M. LAGWAL1, N. BHARDWAJ2, “A Survey on Load Balancing Methods and Algorithms in Cloud Computing”, International Journal of Computer Sciences and Engineering Open Access, Section: Survey Paper, Product Type: Journal Paper Volume-5, Issue-4, Page no. 46-

51, Apr-2017. (Published)

2. MONIKA LAGWAL, NEHA BHARDWAJ, “ Load balancing in Cloud Computing using Genetic Algorithm”, International Conference on Intelligent Computing and Control Systems ICICCS 2017, 978-1-538627457/17/$31.00 ©2017 IEEE. (Published)

REFERENCES

[1] Sakshi, Navtej Singh Ghumman “CLOUD COMPUTING MODEL AND ITS LOAD BALANCING ALGORTIHMS” 2016 International Conference on Computing for Sustainable Global Development (INDIACom).

[2] Jayashree Agarkhed, Ashalatha Rb, ”Dynamic Load Balancing Methods for Resource Optimization in Cloud Computing Environment” IEEE INDICON 2015 1570166709.

[3] Shikha Garg, Dr. D.V. Gupta and Dr. Rakesh Kumar Dwivedi, “Enhanced Active Monitoring Load Balancing Algorithm for Virtual Machines in Cloud Computing” 2016 ISBN: 978-1-5090-3543-4, Proceedings of the SMART -2016, IEEE Conference ID: 39669 5th International Conference on System Modeling & Advancement in Research Trends, 25th_27'h November, 2016 College of Computing Sciences & Information Technology, Teerthanker Mahaveer University, Moradabad, India.

[4] Ronak R Patel, Swachil J Patel, “IMPROVED GA USING POPULATION REDUCTION FOR LOAD BALANCING IN CLOUD COMPUTING”,2016 Intl. Conference on Advances in Computing, Communications and Informatics (ICACCI), Sept. 21¬24, 2016, Jaipur, India IEEE.

[5] Reena Panwar, Bhawna Mallick “Comparative Study of Load Balancing Algorithms in Cloud Computing” International Journal of Computer Applications (0975 – 8887) Volume 117 – No. 24, May 2015.

[7] Shang-Liang Chen , Yun-Yao Chen , Suang-Hong Kuo,”CLB: A novel load balancing architecture and algorithm for cloud services”, m3Gsc; February 25, 2016;19:24 ].

[8] Aarti Singha, Dimple Junejab, Manisha Malhotraa,” Autonomous Agent Based Load Balancing Algorithm in Cloud Computing”, International Conference on Advanced Computing Technologies and Applications (ICACTA- 2015).

[9] Hussain A Makasarwala, Prasun Hazari, “Using Genetic Algorithm for Load Balancing in Cloud Computing”, 978-1-5090-2047-8/16©2016 IEEE.

[10] D. Saranya, L. Sankara Maheswari, “Load Balancing Algorithms in Cloud Computing: A Review”, ISSN: 2277 128X/Volume 5, Issue 7, July 2015.

[11] Ronak R Patel,Swachil J Patel, “IMPROVED GA USING POPULATION REDUCTION FOR LOAD BALANCING IN CLOUD COMPUTING”,2016 Intl. Conference on Advances in Computing, Communications and Informatics (ICACCI), Sept. 21¬24, 2016, Jaipur, India IEEE.

[12] Shikha Garg', Dr. D.V. Gupta2 and Dr. Rakesh Kumar Dwivedi, “Enhanced Active Monitoring Load Balancing Algorithm for Virtual Machines in Cloud Computing” 2016 ISBN: 978-1-5090-3543-4, Proceedings of the SMART -2016, IEEE Conference ID: 39669.

[13] Awatif RagmaniC , Amina EI Omri, Noreddine Abghour, Khalid Moussaid, Mohammed Rida “A Performed Load Balancing Algorithm for Public Cloud Computing U sing Ant Colony Optimization” 978-1-4673-8894-8/16/$31.00 ©2016 IEEE.

[14] Mr. M.Ajit, Ms. G. Vidya, “VM Level Load Balancing in Cloud Environment”, 4th ICCCNT 2013 July 4-6, 2013, Tiruchengode, India.

[15] Pavan Gurve, Aishwarya Mishra, “An Approach for the Load Balancing in Cloud Based on the Dynamic Threshold”, International Journal of Science and Research (IJSR)ISSN (Online): 2319-7064 Volume 3 Issue 10, October 2014.

Thank you