understand the setting of video-based service provided through the cloud framework.
By sampling the service in practice, we also conduct an online experiment for understanding the effect provided through a heuristic mechanism of load balancing.
The theoretical study shows that the service quality of the video-based service depends on the CPU throughout and the bandwidth of network regarding the cloud infrastructure. Intuitively, it comes to mind that the CPU throughout of the server and the client as well as the bandwidth of the network dominate the service quality. The Equation (4) in Section 4 (quantum+max(packagetranstime+delaytime, (BOUNG-1)*
quantum) < playtime) confirms this relationship. It explicitly specifies how the CPU throughout and the bandwidth affect the service quality. Specifically, quantum (the time quantum) and playtime (the time of playing packages generated in a time quantum) both are determined based on the CPU throughout. And packagetranstime (the normal transmission time of packages generated in a time quantum) and delaytime (the transmission delay time) depend on the bandwidth. For different system and network environments, the constraint shows how we adjust the setting to provide smooth video services. If the values of quantum, packagetranstime and delaytime are larger (poor network transmission), we should set the BOUND smaller (so that each VM handles less connections) to avoid potential violation against the service quality. On the other hand, if packagetranstime is larger (a sequence can be displayed a long time on the client side), it is possible that we can set the BOUND larger to allow a VM dealing with more connections simultaneously. Another interesting finding is that theoretically, we can estimate the proper settings to satisfy the service quality given the performance of the computer and network environment (by observing systematical measurement), and hence can be applied to different system environment and architecture. The value of system reliability also can calculate from this equation.
In this paper, we propose that the backstage of the video service – the iPalace Video Channel is under the two phase service system architecture. The service system regarding the phase I is a multichannel, single-phase service system with a variant amount of VMs such that every arrival is served by a VM and there is no balking, and is responsible for receiving the user initial request and responding with the index webpage which directs the user to the corresponding VM. The service system regarding the phase I is a M/D/1: //FCFS queuing system and is designed to satisfy the service quality and accomplish the load balancing mechanism. The service system
‧
responsible for setting up the connection between the User and the corresponding VM and providing the subsequent video service. Defining the type of the service system helps us systematically formulate the relationship between the service quality and the CPU throughout of the server and the client as well as the bandwidth of the network.The two phase service system is designed especially for video-based services because the traditional architecture of the text-based service system is not suitable to the video-based service. Enterprises who intend to provide video services could adopt our queuing model to control the service quality more validly. However, our load balancing strategy is applied to the datacenter in single location. In the case of datacenters in multi-location, enterprises could combine geographic load balancing strategy and our queuing model.
We present the load balance mechanism on the premise that satisfying the service quality. We set up a VM - the Monitor which periodically detects the number of arrivals served by each running VM, selects the VM that has the minimal number of arrivals as the corresponding VM and the next coming arrival(s) will be directed to this VM. On the other hand, if it exceeds the bound, we will increase the size of the cluster by initiating a new VM to join the cluster and dispatch the next arrival to this VM. In the end of monitoring procedure, the unemployed VMs will be released to an idle VM pool. Another VM – the Dispatcher dispatches the arrivals to the corresponding VM, and initiates a new VM and selects it as the corresponding VM when the number of arrivals that the corresponding VM handles with counted by the Dispatcher equals the BOUND.
We conduct an experiment by sampling the video-based service in practice and evaluating the presented approach online. In the experiment we observe that the peak load does not accumulate by a certain VM in the cluster and none of the peak load lasts on a specific VM for a long period of time. The load balance mechanism shows its effectiveness in this setting, but we have to integrate the load balance mechanism with the setting of an upper bound of number of arrivals to be served by a single VM to avoid the peak to be large enough to deteriorate the service quality. We also observe that there are fewer click-and-responses among the server and the user in the video-based service compared to the text-based webpage services.
Most of the researches of load balancing do not take the service quality into consideration, but focus on the performance of the load balance mechanism by conducting experiment or theoretically inferring. Except theoretically inferring and conducting experiment, we also apply queuing models into the load balancing mechanism and try to keep the service quality in high standard in this paper.
‧
Adler, B. 2010. “Load balancing in the cloud: Tools, tips, and techniques,” Technical report, RightScale, Inc..
Adhikari, V. K., Jain, S., Zhang, Z. 2010. “YouTube traffic dynamics and its interplay with a tier-1 ISP: An ISP perspective,” Proceedings of the 10th annual conference on Internet measurement, pp. 431-443.
Amazon Web Service, available at http://aws.amazon.com/, accessed March 23rd, 2012.
Amazon ELB, available online at http://aws.amazon.com/elasticloadbalancing/, accessed March 23rd, 2012.
Aicache Web Accelerator, available online at http://aicache.com/, accessed March 23rd, 2012.
Adan, I., Resing, J. 2001. Queueing Theory. Eindhoven, The Netherlands, Department of Mathematics and Computing Science, Eindhoven University of Technology.
Bilderbeek R., Hertog P., and Marklund G. 1998. Services in innovation: Knowledge intensive business services (KIBS) as co-producers of innovation, 2000, STEP group, SI4S report, 3.
Cherkasova, L. 2000. “FLEX: Load Balancing and Management Strategy for Scalable Web Hosting Service,” IEEE Symposium on Computers and Communications, pp.
8 - 13.
Christodorescu, M., Sailer, R., Schales, D., Sgandurra, D., Zamboni, D. 2009. “Cloud security is not (just) virtualization security: a short paper,” Proceedings of the 2009 ACM workshop on Cloud computing security, pp. 97-102.
Endrei, M., Ang, J., Arsanjani, A., Chua, S., Comte, P., Krogdahl, P., Luo, M., Newling, T. 2004. “Patterns: service-oriented architecture and web services,” IBM Corporation, International Technical Support Organization.
Fielding, R., Gettys, J., Mogul, J., Frystyk, H., Masinter, L., Leach, P. 1999.
Berners-Lee, T.. “Hypertext transfer protocol – http/1.1”, IETF RFC 2616.
Google AppEngine, available online at
https://developers.google.com/appengine/?hl=zh-TW, accessed March 3rd, 2012.
HAProxy, available online at http://haproxy.1wt.eu/, accessed March 3rd, 2012.
Hoffman, D. L. and Novak, T. P. “Marketing in Hypermedia Computer-Mediated Environment Conceptual Foundations,” Journal of Marketing, Vol. 60, 1996, pp.
50-68.
Ishii, A. and Suzumura T. 2011. “Elastic stream computing with clouds,” IEEE International Conference on Cloud Computing, pages 195–202.
‧
Johnson, S. P., Menor, L. J., Roth, A.V., and Chase, R. B. 2000. “A Critical Evaluation of the New Service Development Process,” New Service Development, Thousand Oaks, Calif, pp.18.
Krishnan, R., Madhyastha, H. V., Srinivasan, S., Jain, S., Krishnamurthy, A., Anderson, T., Gao, J. 2009. “Moving Beyond End-to-End Path Information to Optimize CDN Performance,” Proceedings of the 9th ACM SIGCOMM conference on Internet measurement conference IMC, pp. 190-201.
Liu, H. and Wee, S. 2009. “Web Server Farm in the Cloud: Performance Evaluation and Dynamic Architecture,” Cloud Computing, pp. 369-380.
Lipsky, L. R. 2009. “Queuing theory: A linear algebraic approach,” Springer Verlag, pp. 28.
Microsoft Azure, available online at http://www.windowsazure.com/zh-tw/, accessed March 3rd, 2012.
Maglio, P.P. and Spohrer, J. 2008. “Fundamentals of service science,” Journal of the Academy of Marketing Science (36:1), pp. 18-20.
Moreno-Vozmediano, R., Montero, R. S., Llorente, I. M. 2011. “Elastic management of web server clusters on distributed virtual infrastructures,” Concurrency and Computation: Practice and Experience (23:13), pp. 1474-1490.
Namjoshi, J and Gupte, A. 2009. “Service Oriented Architecture for Cloud based Travel Reservation Software as a Service,” Cloud Computing, pp. 147-150.
National Palace Museum, available online at
http://tech2.npm.gov.tw/da/eng/active_news_1.html , accessed May 3rd, 2012.
Nobel , C. 2011. “A New Model for Business: The Museum,” accessed at 12/28/2011 via http://hbswk.hbs.edu/item/6770.html.
Phifer, L. 2006. “Establishing a Geographically-Distributed Internet,” Technical report, Coyote Point Systems, Inc., available at http://aws.amazon.com/, accessed March 23rd, 2012.
Piórkowski, A., Kempny, A., Hajduk, A., Strzelczyk, J. 2010. “Load Balancing for Heterogeneous Web Servers,” Communications in Computer and Information Science vol. 79, pp. 189-198.
Patterson, D., Armbrust, M., Fox, A., Griffith, R., Jseph, A. D., Katz, R. H., Kownwinski, A., Lee, G., Rabkin, A., Stoica, I. 2009. “Above the clouds: A Berkeley view of cloud computing,” Tech. Rep. No. UCB/EECS-2009-28.
Berkeley, CA: University of California.
Rasch, P. J. 1970. “A queueing theory study of round-robin scheduling of time-shared computer systems,” Journal of the ACM (JACM) (17:1), pp. 131-145.
Research Development and Evaluation Commission 2011, available online at http://www.rdec.gov.tw/mp110.htm, accessed March 23rd, 2012.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
43
Silberschatz, A., Gagne, G., Galvin, P. B. (2004) “Operating System Concept 7th edition,” Addison-Wesley.
Sakata, M., Noguchi, S., Oizumi, J. 1971. “An analysis of the M/G/1 queue under round-robin scheduling,” Operations research, pp. 371-385.
Turner, M., Budgen, D., Brereton, P. 2003. “Turning software into a service,”
Computer, (36: 10). pp. 38-44.
Vargo, S. L. and M. A. Akaka. 2009. “Service-dominant logic as a foundation for service science: clarifications,” Service Science, (1:1). Pp. 32-41.
Wee, S. and Liu, H. 2010. “Client-side Load Balancer using Cloud,” Proceeding SAC '10 Proceedings of the 2010 ACM Symposium on Applied Computing, pp. 399-405.
Wolff, R. W. 1982. “Poisson arrivals see time averages,” Operations research (30:2), pp. 223-231.
Youtube, available online at http://www.youtube.com/, accessed March 23rd, 2012.