2012 26th International Conference on Advanced Information Networking and Applications Workshops
An Exploration of The Optimization of Excutive Scheduling in The Cloud Computing
Department of Information Management St. John’s University New Taipei City, TAIWAN, ROC firstname.lastname@example.org
Department of Information Management St. John’s University New Taipei City, TAIWAN, ROC email@example.com
Abstract-The cloud computing is roughly in Iaas, PaaS, SaaS these 3 categories, but they all have a common principles which are to use pays principle of how much is how many. It is not the same way as traditional building to purchase the largest estimates of the number of users’ hardware resources. Therefore, the cloud computing also has environmental protection significance. However, how effective and full use of maximum resources in the cloud computing is a desirable subject for discussion. The study suggests use of the working scheduling models to improve the resource allocation in the cloud computing, however, the cloud computing and distribution of resources is more responsive application needs to be time and resources optimal applications. We analog users in accordance with demand category of work parameters, and the scheduling parameters to be access to Internet usage or take the single time to do the parameter, by means of queuing theory for scheduling models such as the demand, scheduling the parameters into model, generate reference data, through the analysis of data to compare the various scheduling characteristics. The future of the cloud computing systems to be a frame of reference, and build their own cloud computing. Keywords- Cloud Computing, Queuing Theory, Network Services
A Human Resources Management System (HRMS) or Human Resources Information System (HRIS), refers to the systems and processes at the intersection betweenhuman resource management (HRM) and information technology. It merges HRM as a discipline and in particular its basic HR activities and processes with the information technology field, whereas the programming of data processing systems evolved into ...
information provider faces many problems, such as hardware construction, and the Internet traffic estimates, and the Internet building is associated problems with the users. On the assumption by conventional means to solve these problems, several like Amazon, Yahoo had the same sufficient manpower and financial resources. As a result, the Internet world with the actual society is like the state that cannot afford to buy storefront to rent shop, cannot afford to buy host to rent hosts, and it is an environment to bring up today’s rise of the cloud computing. A. Research motivation In recent years, the Google, Microsoft, Amazon large companies such as efforts to promote the cloud computing has become the focus IT industry, causing global effects, domestic and foreign companies invested the cloud computing research and development. Literature shows that the use of the cloud computing can reduce wastage of resources and efficient upgrade effectiveness, increasing the proper application of business computer resources or improvements computing device’s effectiveness and efficiency of the cloud computing itself can achieve the best applications. It also can import working scheduling model for best use rate of hosts.
B. Research purpose This study purposes are as the followings: (1) Explore the difference of the working scheduling in the cloud computing. (2) Explore the working scheduling applications in the cloud computing. (3) As a result of research and analysis provide to relevant units for reference in the future research, and the contribution to practical area. II. LITERATURE A. The Cloud computing architecture The rise in the cloud computing broad the attention of many systems and services to use the cloud computing to save resources or sharing resources, and the the cloud computing being better than new technologies rather than call it a concept of evolution, because the cloud computing successor distributed computing and grid computing between the two big concept . The cloud computing is still in the development stage in the no standard definition and specifications and so it is necessary for some time to define the field of the cloud applications. The cloud computing can solve some resource sharing. The cloud computing needs to master virtual technology and
Family Village For Families of Children with Special Needs – http://www.familyvillage.wisc.edu/websites.html This website offers a wide variety of resources for the community to use when trying to navigate around the internet. It has sources such as Parent to Parent Matching, Mental health Websites, Diagnosis, and Internet resources. The Tulsa Area Alliance on Disabilities – ...
The cloud computing in recent years were widely discussed and apply. The cloud computing features in addition to via the Internet users access to resources for application, another key features is to use resources even cost, the users a surfeit of waste and sank costs, and that many emerging Internet services while in the establishment of hardware change the traditional purchase turn to lease to be estimated not be easy hardware of costs. The movie sharing site Animoto is a good example, using Amazon EC 2 platform  to deal with three days soared 10 times the number of users. Another classic case is one familiar Twitter site, it is to use Amazon S3  as storage equipment to deal with growing use of the case of infrastructure as a service (IaaS) based structure to establish the software as a service (SaaS) . The rise in the cloud computing, network speed and the Internet information on the Internet volume are inseparable as Hill. If the network speed of unhappiness or information on the Internet low level, one can let the cloud computing becomes rampant of the information technology sector new favorites. As a result of the Internet ever-increasing demands, it gradually become accustomed to the information on the Internet’s rapidly 10,000 change, and thus the requirement of the Internet is rising. Therefore, let
978-0-7695-4652-0/12 $26.00 © 2012 IEEE DOI 10.1109/WAINA.2012.159
service-oriented technical two important technical and can effectively perform the advantages . Fig. 1 shows the cloud structure, described in the next three sections.
Scheduling model Queuing theory proposed by mathematician A. k Erlang as early as in 1909, through mathematical modeling describe waiting on system. The main framework can be divided into four main elements, inputs, service equipment, system capacity, service rules. Simple to say, it is amount of the need to be services will produce queuing situation, daily life often see such a situation, such as petrol filling station refueling, bank deposit money, customs inspection. The model is many, we select suitable for use in scheduling the models to study and explore . M/M1 M/M/1 is the most basic waiting models, only a single services platform scheduling the Pu Watts Son distribution, no special note takes with the rule of FCFS, scheduling lead time and system services for index distribution . The import M/M/1 model by the systems should be the service rateρ(1) or a single scheduling of W (4) as the comparison and adjustments based on, the current study to select the system of service rates on the basis. Formula: ρ= λ/μ (1) (2) Po=1-ρ ; L=ρ/1-ρ (3) W=1/λ*L (4) M/MS M/M/S is M/M/1 further, the two greatest differences between the number of service platform differences, no special note takes with the rule of FCFS, scheduling lead time and system services for index distribution. The import M/M/S model by the systems should be the service rateρ(5) or a single scheduling of W (9) as the comparison and adjustments based on, the current study to select the system of service rates on the basis. Formula: ρ= λ/Sμ (5) Po=1/( +1/S (1-ρ) Lq=ρ/s (1-ρ) ( L=Lq+λ/μ W= L/λ
COMPARATIVE STUDIES OF NATIONAL CIVIL SERVICE SYSTEMS by NEIL M. CAN LOBO Laguna College of Business and Arts Paper prepared for presentation at Civil Service Systems in Comparative Perspective, School of Public and Environmental Affairs, Indiana University, Bloomington, Indiana, April 5-8, 1997. ABSTRACT The Philippine civil service system is a legacy of the country's colonial experience, first ...
Figure 1. Framework of cloud computing
Infrastructure services Infrastructure as a service (IaaS) is the infrastructure of PaaS and SaaS. It provides a framework mainly provide hardware resources and environment to users easy to use and is the basic facilities to provide services, such as Amazon EC 2 offer Virtual Machine let developers can install the OS with their planning as well as required software, and that the three services the freest . Platform service Platform as a service (PaaS) is within the framework of above IaaS and provide a complete program development platform and operating system platform to writing. The developers through the Internet software and services provided to the general users can implement programs, such as Google app engine . PaaS with the biggest difference between IaaS in ductility, while the program going up the burden, the server can automatically extend to cope with extra burden; however, IaaS cannot achieve automatically extended. The IaaS developer need to consider and judgment extended to additional resources. Software services Software as a service (SaaS) is to use the mode of service provides software, software services provided through the Internet software function service model, in addition to simple use, it can also let users’ use and development cost reduction.
Introduction Grading and reporting are relatively recent phenomena in education. Throughout much of the nineteenth century most schools grouped students of all ages and backgrounds together with one teacher in one-room schoolhouses, and few students went beyond college studies. The teacher reported students learning progress orally to parents, usually during visits of parents in their schools. ...
Simple to say, users are not necessary to install software, and that they do not have the burden with the relative to the hardware. Software running on major suppliers or a third party, the Internet access doesn’t have the installation of the Office can also use Word or Excel Spreadsheet editing and do it. B. The cloud computing category On the current generally agreed to the cloud which is divided into three categories, namely the public, private cloud, mixed cloud. Public cloud was enterprises and consumers joint use of the cloud environment, a number of different users use in the cloud suppliers to let users use common resources in the cloud suppliers . Within the enterprise private cloud of the cloud environment, the users for the enterprises or internal organization members use the common cloud resources, the business outside people cannot use the cloud services provided by resources . The mixed cloud is mixed public cloud and private cloud. For someone using the planning, mostly companies have a certain amount of IT infrastructure, through a small amount of investment system upgrade, it can enjoy the benefits of the cloud. Generally, the safety and reliability applications on relatively low are in the public cloud, private cloud part needs security information . z z
(6) (7) (8) (9)
III. RESEARCH APPROACH A. System development process
Figure 2. System development process chart
System development process as shown in Fig. 2, we make needs analysis first, through model set doing the
system design and system put in place to give rise to a data to further discuss model features. B. System structure Architecture as shown in Fig. 3: (1) Through VM system of many server hosts, the simulation of multiple hosts’ status, setup models of the state environment (2) Through server analog moves, process scheduling and record the time of data in order to later analysis.
1.1 discuss the characteristics of food production and food and beverage service systems Food production: systems eg traditional, batch cooking, call-order, centralised, assembly kitchens, sous-vide, cook-chill, cook-freeze Service: systems eg table service, counter service, a la carte, table d’hote, silver service, family service, plate service, gueridon service, specialist food service systems ...
TABLE II. KENDALL COMMON TYPES AND CODE
Item M D G Ek
Explanation Index distribution, eg: M/M/1 Certainty distribution, eg:M/D/1 The general distribution, eg: M/G/1 Erlang-k distribution, eg: G/Ek/1
TABLE III. PENDING THEORETICAL EXPLAIN
Parameter n ρ
Figure 3 Systematic structures
Model set In daily life, we can often see waiting in line, let the people to associate in the cloud computing is through queuing theory to address the resource allocation. The reason of selected M/M/1 and M/M/S the two models were major they are one of the most basic model of queuing theory, and M/M/S is M/M/1 extension made easily understand, and convenient to compare and analysis. So first define the platform scheduling =10, 20, 30, by means of the changes of single average scheduling execution time to observe the change of system service rates ρ(5), and find the time to be suitable for increasing the service platform. M/M/1/FCFS/ / model This is a basic waiting mode, only a single services platform scheduling the Pu Watts Son distribution, the system capacity and scheduling quantities are set to infinity, at the same time assumed that system scheduling no time out set, no matter how long waiting will wait until implementation of the completion . z M/M/S/FCFS/ / model In the system, scheduling arrived at the time of system and system service time for index distribution, and the service platform of S, and to introduce FCFS . Queuing theoretical terminology and code explain Queuing theory terminology and code explained the Kendall mark as TABLE I, TABLE II and TABLE III.
TABLE I. KENDALL MARK STATEMENT 
Effective reached rate The average rate of system services
Inflation Rates By: Anonymous The price of one currency in terms of another is called the exchange rate. The exchange rate affects the economy in our daily lives because it affects the price of domestically produced goods sold abroad and the cost of foreign goods bought domestically. "Mexicans use pesos, French use francs, Austrians use schillings, and this use of different monies by different ...
μ S L W
Service rate of the service stations Services platform quantity There are many scheduling of average platform Single scheduling implementation time
D. System simulation (1) The system note: this system to adopt Java programming language written, VM use virtual host simulate multiple hosts, through scheduling server simulation test records. (2) Architecture features: as shown in Fig. 4 users choose modalities for the implementation of the data will choose the paradigm is different.
Position A B C D E F
A/B/C/D/E/F Explanation Intervals arrived at the distribution rates The chances of distribution service time The number of service platform Queuing (waiting) rules (FCFS: First Come First Service) The largest number of system hosts (awaiting number + excuting number) Scale quantity of the resource
Figure 4 Systematic use case
In M/M/1 models simulation experiment, we use the system services rateρ(1) as the vertical coordinates, average length of service for X-coordinate, the change in our average length of service to indicate an average reached rate =10, 20, 30 into M/M/1 models of change of the system service rates ρas shown in Fig. 5.
Figure 5. M/M/1 system service rate change
Figure 8. M/M/1 and M/M/2
Moreover, in M/M/2 models simulation experiment, we use the system service rates ρ(5) as the vertical coordinates, average length of service for X-coordinate, the change in our average length of service to indicate an average reached rate =10, 20, 30 brought into M/M/2 model of change of system service ratesρas shown in Fig. 6.
By the phenomenon, we can see that we can use of a service rates as a platform basis for an increase, reduce unnecessary system waste of resources. IV. CONCLUSION In this article, we have the cloud computing and the queuing theory on the basis and the simulation of users in accordance with demand category of parameters. In addition, scheduling the parameters can access to Internet usage or a singlet the time to do the parameters. Use the cloud computing through queuing theoretical models of the produced data that try to classify the best of the model, to provide an effective feasibility of proposals to help resolve the cloud computing user could provide a basis, and achieve more closely user’s computer resource requirements. REFERENCES
Lin, Chi-jen, Management mathematics, Koritsu Books, Taipei, 2006.  Chen, Ying, The cloud strategy, the world magazine, Taipei, 2010.  Yip, Wai-CHEUNG, Management mathematics, 5 South books, Taipei, 1999.  Yang, Che, The cloud computing technical guidelines, Okahata asset management, Taipei , 2010.  Yeung, Kam-cheung, and Chen, Teng-source, The management mathematics missile theory, double-ye book Gallery, Taipei, 2009.  Deng, Wei Don, The BP social publications, the clouds computing have a bowel movement secret, city-state cultural, Taipei, 2010.  Amazon Elastic Compute Cloud – EC2.http://aws.amazon.com/ec2/.  Liang-Jie Zhang and Qun Zhou, “CCOA: Cloud Computing Open Architecture,” 2009 IEEE International Conference on Web Services, 2009, pp. 607-616.  Luqun Li, “An Optimistic Differentiated Service Job Scheduling System for Cloud Computing Service Users and Providers,” Third International Conference on Multimedia and Ubiquitous Engineering, 2009.  Robert L, Yunhong Gu, On the Varieties of Clouds for Data Intensive Computing, 2009.  Michael Armbrust, Armando Fox, Rean Griffith, Anthony D. Joseph, Randy Katz,Andy Konwinski, Gunho Lee, David Patterson, Ariel Rabkin, Ion Stoica, and Matei Zaharia, Above the Clouds: A Berkeley View of Cloud Computing, 2009. 
Figure 6. M/M/2 system service rate change
as a result of the simulation experiments, the figure 5 and 6 of =20 removed to compare M/M/1 and M/M/2 of the system service ratesρas shown in Fig. 7, through the system services rate ρ(5) can see M/M/2 better than M/M/1 in system performance, we also can use single scheduling of W (9) to see the differences.
Figure 7. Comparisons between M/M/1 and M/M/2
We have queuing theory was informed that the higher of system service rates ρ(5), the lower of system idle rate Po(6), therefore, the service rates ρ=1 as the maximize effectiveness, ifρ> 1 there will be some scheduling unable to deal with the rate would be the demand greater than services. So, we can start using M/M/1 the framework, whileρ=1 one additional services platform import M/M/2 framework, infer from this, while we are in ρ=1 increase one service platform, the Fig. 8 can be seen from the increase in service rates would be one of the service platform and decline.