Sei sulla pagina 1di 6

IPASJ International Journal of Computer Science (IIJCS)

A Publisher for Research Motivation ........

Volume 1, Issue 6, December 2013

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm Email: editoriijcs@ipasj.org ISSN 2321-5992

Qos Aware and Cost Based Optimization of Service Composition Using Genetic Algorithm
Mrs.K.Rajasri, R.Arundurai and K.Dileepan Ady
Senior Assistant Professor, CSE, Christ College of Engg&Tech, Pondicherry, India Student, CSE, Christ College of Engg&Tech, Pondicherry, India Student, CSE, Christ College of Engg&Tech, Pondicherry, India

ABSTRACT
In cloud computing to improve the QoS (Quality of Service) between cloud server and the client and it also to be check the possibilities to handle the SLA violation problem. In our approach how to handle SLA violation problem in cloud services and predict it. In the existing work to use the runtime adaptation tool and PREVENT framework to reach the SLA violations. In this paper, we are using the PREVENT (prediction and prevention) framework. To use this framework to prevent and how to handle the SLA violation problem. This approach to improve QoS (Quality of Service) for user and cloud server and it also reduces maximum cost for the user to access this service. To use different algorithm to improve the QoS. This approach it also to reduce the time to access and eliminate lagging of data, it is also provide priority based service to the user. In this approach display the result of SLA violation and QoS depending on the different algorithm to show the result for high or low.

Keywords: Service-level agreements, service-level objectives, QoS, cloud service composition, heuristic algorithm, genetic algorithm, adaptation, optimization.

1. INTRODUCTION
Service-based applications have seen tremendous research activity in the last years, with many important results being generated around the world. This global interest is justified by the ever increasing services industry, which is still only starting to explore the potential that new paradigms like Everything-as-a-Service (XaaS) or cloud computing provide. However, to fully realize this potential, research and industry alike need to focus more strongly on nonfunctional properties and quality issue of services (generally referred to as QoS). In the business world, QoS promises are typically defined within legally binding service-level agreements (SLAs) between clients and service providers, represented, e.g., using WSLA. SLAs contain service-level objectives (SLOs), i.e., concrete numerical QoS objectives, which the service needs to fulfill. a. Cloud Computing Cloud computing it means it is allocates the resource to the users. Cloud computing is the next stage in evolution of the Internet. Cloud computing provides the means through which everything from computing power to computing infrastructure, applications, business processes to personal collaboration can be delivered to you as a service wherever and whenever you need. Cloud computing is the use of computing resources are can be delivered as a service over a network (typically the Internet). The name comes from the use of a cloud-shaped. To use cloud service in all services with a user data value.
Client Server

Figure.1.1. Cloud logical diagram

Volume 1 Issue 6 December 2013

Page 24

IPASJ International Journal of Computer Science (IIJCS)


A Publisher for Research Motivation ........

Volume 1, Issue 6, December 2013

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm Email: editoriijcs@ipasj.org ISSN 2321-5992

Cloud computing it can be developed by Google, now days the Google it to be use and to be achieve the Google cloud developer. The cloud computing it is very easy method that it is to be used, a small application it can also to be download and used by the cloud. To use the cloud to access the cloud server anywhere to connect it. This domain it should always to be a various model to be used. The user to buy a cloud server and to select any one service to be connects that service and access that service anywhere. Users can store the files and notes to be store in the cloud server. Cloud server is dividing into public sector and private sector. In each of them the cloud server it should be work and different type of security system is available. To use any one to develop their work in the cloud server.

Operating System

APP APP

Operating System

APP APP APP APP

Operating System

Virtualization Layer

Figure.1.2. Working system of cloud server

2. EXISTING SYSTEM
In the existing system to formalize the SLA violation problem and it to be reduce the maximum cost for the give predicted SLA cost. In existing to reduce the maximum cost to use cost optimization method. To use the possible algorithms to solve the SLA violations and optimization problem. In existing system to use genetic algorithm to find the SLA violation problem and to use branch and bound algorithm for cost optimization problem. An existing system to produce the results is shows different between the various cost value and different services for use different algorithms. The PREVENT framework it is consists of some steps monitor, analyze, plan and execute those are in the vision of autonomic computing. To use branch and bound algorithm to generate the optimization formula for reduce the maximum cost in the total cost arise. In this to use adaptation action of service composition to define the specify instances of SLA value and to choose the constraint to use it. In this to use more complex algorithms to perform the adaptation action. 2.1 Genetic Algorithm A Genetic Algorithm it is the natural way of search process for the all evolution process. Mostly the genetic algorithm is used by to solve the optimization problem by using the any other natural evolution. The genetic algorithm it is independent to the larger of evolutionary algorithm (EA), and it is also to be solve the optimization problem. In the genetic algorithms are search methods which were developed by John Holland and are based on the principles of natural selection, a biological process in which stronger individuals are likely the winners in a competitive environment. GAs are usually used for producing acceptable solutions to optimization problem. To use genetic algorithm in the optimization method, to first create a simple genetic search and it to be check the optimization problem. Using this type of the method the problem solution it must to satisfied and domain systems are undergone by a single algorithm. In the existing work the genetic algorithm it must to be used in the situation, to find the SLA violation and how to handle it. That cases we can be develop the various form or methods to be choose in the genetic search, then the search is to very efficient and that can be resolved. To be improving Qos (Quality of Service) in cloud server it can be easy to be use the genetic search, but sometimes it not be improve the service and to be change it. When to be use the genetic algorithm it to be check the fitness function and to be check the every individual population of that service to be evaluate, after perform the genetic search.

Volume 1 Issue 6 December 2013

Page 25

IPASJ International Journal of Computer Science (IIJCS)


A Publisher for Research Motivation ........

Volume 1, Issue 6, December 2013

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm Email: editoriijcs@ipasj.org ISSN 2321-5992

After it to be check the fitness function value then it to be initiated the population values and to perform the genetic search. While identified or to be check the fitness function value, in very few cases do not define the fitness function value. The genetic algorithm it is to be used after each successive value of genetic search the population it must to be generated by the new generation method to be created. To be use the some bit value it can to be representing fitness object in the searching process. In the bit process method it can to be choose different types of objects in the genetic search method. The general pseudo code for Genetic Algorithm, 1. Choose initial population 2. evaluate each individuals fitness 3. determine populations average fitness 4. repeat 5. select best-ranking individuals to reproduce 6. mate pairs at random 7. apply crossover operator 8. apply mutation operator 9. evaluate each individuals fitness 10. determine populations average fitness 11. until terminating condition (e.g. until at least one individual has generations have passed). 2.2 Branch and Bound Algorithm Branch-and-bound is a very general deterministic algorithm for solving optimization problems. The high-level idea of this approach is to enumerate the solution space in a smart way so that at least some suboptimal solutions can be identified and discarded prematurely. In our Branch-and-Bound approach, we prune a partial solution in two cases, if the partial solution already contains at least one conflict, or if the partial solution already prevents all SLA violations without applying any more action. The branch and bound method it to be applied for mostly in the optimization method. It is very fast method to be solved optimization problem compared to the greedy search method. The branch and bound method it that is similar to the BFS method, it can be generate the sub nodes. Some cases the time complexity is too low in branch and bound technique, in that cases we are to be use the various optimal solution to satisfy the problem. The all branch and bound technique it to be generate the some optimal solution and to use the nodes to be expand it the process. The general pseudo code for Branch and Bound Algorithm 1. Search (t, r, best) 2. pre: t=solution space tree 3. r=vertex in t 4. best= best solution found so far 5. post: best= best solution found after searching sub tree rooted at r 6. if r is a complete solution more optimum than best then set best=r 7. generate the children of r 8. compute bounds for vertices in sub trees of children 9. v1,vk=feasible children with good lower bounds 10. for i:=1 to k 11. if vi has a promising upper bound 12. then search (t, vi, best) 2.3 Drawbacks in Existing System Service cost is more. More time and lagging of data. Problem in the predicted SLO value. The Quality of Service is low, because of SLA problem.

the desired fitness or enough

Volume 1 Issue 6 December 2013

Page 26

IPASJ International Journal of Computer Science (IIJCS)


A Publisher for Research Motivation ........

Volume 1, Issue 6, December 2013 3. PROPOSED SYSTEM

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm Email: editoriijcs@ipasj.org ISSN 2321-5992

In our approach to provide the better Qos (Quality of Service) and to predict the SLA violation problem and to solve the SLA problem. In the proposed system to improve the QoS use some complex algorithms in the system. In the system to develop a PREVENT framework for predict the SLA violation problem. To use heuristic algorithm to for improve the Quality of Service (Qos) between cloud server and client. Our approach it is also provide priority based service and reduce the access time. In the Quality of Service (QoS) it means to adopt the value added services such as response time, throughput, reliability and price. a. Heuristic Algorithm The Heuristic search it is to be developed for solve optimization problem and to improve QoS (Quality of Service). This search it is also to be finding and to prevent SLA violation problem. The heuristic search technique it is very fast and easy solves any problem in quick way, and to generate accurate solution. Some search technique it does not show the accurate value, sometimes it do not show the output value. Some it produce the approximate value or does not produce any value. In the proposed to be improve QoS value between user and the cloud server to be use heuristic search technique, because in this is easy way to generate the all solution and fitness or population value to be generated. To be using this technique in which it is very quickly to be solves it. The heuristic technique is to produce proper solution taken by humans. This technique was accepted by it in the user for good search technique and efficiency it to be checked in the method to be solved easily. The solution of the heuristic search it is to be satisfied the user and developer. Sometimes this technique it wont to be work and to let be used in the any other form of solution method it to be used. The heuristic search technique it is an effort fully form of theories when compared to other theories. To use this heuristic value it can to be measured by in all other solved technique it to be used. To use a heuristic algorithm to be solve a given problem to use some trade off, Optimality Completeness Accuracy and precision Execution time

The general pseudo code for Heuristic Algorithm, 1. S s0; e E(s) // Initial state, energy. 2. sbest s; ebest e // Initial best solution. 3. k 0 // Energy Evaluation count.

4. while k < kmax and e > emax //while time left & not good enough. 5. T temperature (k/kmax) // Temperaure calculation. 6. snew neighbour(s) 7. enewE(snew) // Pick some neighbour

// Compute its energy

8. if P (e, enew, T) > random() then // Should we move to it? 9. s snew; e enew 10. if enew < ebest then // Yes, change state. // Is this a new best? // Save newneighbourtobest found.

11. sbest snew; ebest enew

12. k k+1 // One more evaluation done 13. return sbest // Return the best solution found

Volume 1 Issue 6 December 2013

Page 27

IPASJ International Journal of Computer Science (IIJCS)


A Publisher for Research Motivation ........

Volume 1, Issue 6, December 2013


b. PREVENT Framework

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm Email: editoriijcs@ipasj.org ISSN 2321-5992

Figure. 3.1. Framework for SLA prediction and prevention c. Advantages of Proposed System Reduce the maximum cost. To improve good Quality of Service (QoS). To solve the SLA violation problem. The SLO value problem does not occur any time.

4. CONCLUSION AND FUTURE WORK


For providers of composite web services, it is essential to be able to minimize cases of SLA violations. One possible route to achieve this is to predict at runtime, which instances are in danger of violating SLAs, and to apply various adaptation actions to these instances only. However, it is not trivial to identify which adaptations are the most cost effective way to prevent any violation, or if it is at all possible to prevent a violation in a cost-effective way. Furthermore, we have use both simulated annealing and deterministic local search solution algorithm. We have evaluated these algorithms based on a manufacturing case study and have shown which types of algorithms are better suited for which scenarios. The main current limitation is that adaptation is only considered on instance level that is, for each composition instance separately. We believe that the PREVENT adaptation model can be extended to this kind of SLOs and actions, but new approaches to predict violations and impact models are needed to this end.

References
[1] Philipp Leitner, Waldemar Hummer, SchahramDustdar ,Cost Based Optimization Of Service Compostion Vol., No.2, April-June 2013. [2] A. Dan, D. Davis, R. Kearney, A. Keller, R. King, D. Kuebler, H. Ludwig, M. Polan, M. Spreitzer, and A. Youssef, WebServices on Demand: WSLA-Driven Automated Management,IBM Systems J., vol. 43, no. 1, pp. 136-158, Jan. 2004. [3] L. Bodenstaff, A. Wombacher, M. Reichert, and M.C. Jaeger, Analyzing Impact Factors on Composite Services, Proc. IEEEIntl Conf. Services Computing (SCC 09), pp. 218-226, 2009. [4] B. Wetzstein, P. Leitner, F. Rosenberg, S. Dustdar, and F. Leymann, Identifying Influential Factors of Business Process Performance Using Dependency Analysis, Enterprise Information Systems, vol. 4,no. 3, pp. 1-8, July 2010. [5] P. Leitner, B. Wetzstein, D. Karastoyanova, W. Hummer, S. Dustdar, and F. Leymann, Preventing SLA Violations in Service Compositions Using Aspect-Based Fragment Substitution, Proc.Intl Conf. Service-Oriented Computing (ICSOC 10), 2010. [6] P. Leitner, A. Michlmayr, F. Rosenberg, and S. Dustdar,nMonitoring, Prediction and Prevention of SLA Violations in Composite Services, Proc. IEEE Intl Conf. Web Services (ICWS 10),pp. 369-376, 2010. [7] A. Michlmayr, F. Rosenberg, P. Leitner, and S. Dustdar, End-to-End Support for QoS-Aware Service Selection, Binding, and Mediation in VRESCo, IEEE Trans. Services Computing, vol. 3,no. 3, pp. 193-205, July 2010. [8] L. Xu and B. Jennings, A Cost-Minimizing Service Composition Selection Algorithm Supporting Time-Sensitive Discounts, Proc.IEEE Intl Conf. Services Computing (SCC 10), pp. 402-408, 2010.

Volume 1 Issue 6 December 2013

Page 28

IPASJ International Journal of Computer Science (IIJCS)


A Publisher for Research Motivation ........

Volume 1, Issue 6, December 2013

Web Site: http://www.ipasj.org/IIJCS/IIJCS.htm Email: editoriijcs@ipasj.org ISSN 2321-5992

[9] T. Feo and M. Resende, Greedy Randomized Adaptive Search Procedures, J. Global Optimization, vol. 6, pp. 109-133, 1995. [10] N. Radcliffe and P. Surry, Formal Memetic Algorithms, Evolutionary Computing, vol. 865, pp. 1-16, 1994. [11] R. Jurca, B. Faltings, and W. Binder, Reliable QoS Monitoring Based on Client Feedback, Proc. 16th Intl Conf. World Wide Web (WWW 07), pp. 1003-1012, 2007. [12] L. Zeng, H. Lei, and H. Chang, Monitoring the QoS for Web Services, Proc. Fifth Intl Conf. Service-Oriented Computing (ICSOC07), pp. 132-144, 2007. [13] L. Zeng, B. Benatallah, A.H.H. Ngu, M. Dumas, J. Kalagnanam, and H. Chang, QoS-Aware Middleware for Web Services Composition, IEEE Trans. Software Eng., vol. 30, no. 5, pp. 311-327, May 2004. [14] R. Berbner, M. Spahn, N. Repp, O. Heckmann, and R. Steinmetz, Heuristics for QoS-Aware Web Service Composition, Proc. IEEE Intl Conf. Web Services (ICWS 06), pp. 72-82, 2006. [15] T. Yu, Y. Zhang, and K.-J. Lin, Efficient Algorithms for Web Services Selection with End-to-End QoS Constraints, ACM Trans. the Web, vol. 1, article 6, May 2007. [16] J. Xu and S. Reiff-Marganiec, Towards Heuristic Web Services Composition Using Immune Algorithm, Proc. IEEE Intl Conf. Web Services (ICWS 08), pp. 238-245, 2008. [17] M. Alrifai and T. Risse, Combining Global Optimization with Local Selection for Efficient QoS-Aware Service Composition, Proc. 18th Intl Conf. World Wide Web (WWW 09), pp. 881-890, 2009. [18] F. Rosenberg, M.B. Mu ller, P. Leitner, A. Michlmayr, A. Bouguettaya, and S. Dustdar, Metaheuristic Optimization of Large-Scale QoS-Aware Service Compositions, Proc. IEEE Intl Conf. Services Computing (SCC 10), 2010. [19] T. Unger, F. Leymann, S. Mauchart, and T. Scheibler, Aggregation of Service Level Agreements in the Context of Business Processes, Proc. 12th Intl Enterprise Distributed Object Computing Conf. (EDOC 08), pp. 43-52, 2008. [20] A. Erradi, P. Maheshwari, and V. Tosic, Policy-Driven Middleware for Self-Adaptation of Web Services Compositions, Proc. ACM/IFIP/USENIX Intl Conf. Middleware (Middleware 06), pp. 62-80, 2006.

Volume 1 Issue 6 December 2013

Page 29

Potrebbero piacerti anche