Abstract
Server providers that support e-commerce applications as a service for multiple e-commerce Web sites traditionally use a tiered server architecture. This architecture includes an application tier to process requests for dynamically generated content. How this tier is provisioned can significantly impact a provider's profit margin. In this article we study methods to provision servers in the application serving tier that increase a server provider's profits. First, we examine actual traces of request arrivals to the application tier of an e-commerce site, and show that the arrival process is effectively Poisson. Next, we construct an optimization problem in the context of a set of application servers modeled as M/G/1/PS queueing systems, and derive three simple methods that approximate the allocation that maximizes profits. Simulation results demonstrate that our approximation methods achieve profits that are close to optimal, and are significantly higher than those achieved via simple heuristics.
- Almeida, V., Fonseca, R., Mendes, M. A., and Menasce, D. 2000. Resource management policies for e-commerce servers. Perf. Eval. Review 27, 4 (Mar.). Google ScholarDigital Library
- Challenger, J., Dantzig, P., Iyengar, A., Squillante, M., and Zhang, L. 2004. Efficiently serving dynamic data at highly accessed web sites. IEEE/ACM Trans. Netw. 12, 2. Google ScholarDigital Library
- de Farias, D., King, A., and Squillante, M. 2002. Dynamic control of web server farms. In INFORMS Revenue Management Section Conference.Google Scholar
- Epperson, J. F. 2001. An Introduction to Numerical Methods and Analysis. J. Wiley, New York, NY.Google Scholar
- Federgruen, A. and Groenvelt, H. 1986. The greedy procedure for resource allocation problems: necessary and sufficient conditions for optimality. Oper. Res. 34, 908--918. Google ScholarCross Ref
- Grossglauser, M. and Bolot, J.-C. 1999. On the long range dependence in network traffic. IEEE/ACM Trans. Netw. 7, 5 (Oct.), 629--640. Google ScholarDigital Library
- Harchol-Balter, M., Sigman, K., and Wierman, A. 2002. Understanding the slowdown of large jobs. Perf. Eval. Review 30, 3, 9--11. Google ScholarDigital Library
- Libman, L. and Orda, A. 1999. The designer's perspective to atomic noncooperative networks. IEEE/ACM Trans. Netw. Google ScholarDigital Library
- Liu, Z., Squillante, M., and Wolf, J. 2001a. On maximizing service-level-agreement profits. In Proceedings of the ACM Conference on Electronic Commerce, 213--223. Google ScholarDigital Library
- Liu, Z., Squillante, M., and Wolf, J. 2001b. Optimal control of resource allocation in e-business environments with strict quality-of-service performance guarantees. Tech. rep., IBM Research Division.Google Scholar
- McWherter, D., Schroeder, B., Ailamaki, N., and Harchol-Balter, M. 2004. Priority mechanisms for OLTP and transactional Web applications. In Proceedings of the International Conference on Data Engineering (ICDE 2004). Boston, MA. Google ScholarDigital Library
- Menasce, D., Almeida, V., Riedi, R., Fonseca, R., and Jr., W. M. 2000. In search of invariants for e-business workloads. In Proceedings of the ACM Conference on Electronic Commerce. Minneapolis, MN, 56--65. Google ScholarDigital Library
- Nahum, E. 2002. Deconstructing specweb99. In Proceedings of the WCW'99. Boulder, CO.Google Scholar
- Ryu, B. and Elwalid, A. 1996. The importance of long-range dependence of VBR video traffic in ATM traffic engineering: myths and realities. In Proceedings of the ACM SIGCOMM'96. Palo Alto, CA, 3--14. Google ScholarDigital Library
- Sairamesh, J., Ferguson, D., and Yemini, Y. 1995. An approach to pricing, optimal allocation and quality of service. In Proceedings of the INFOCOM'95, 1111--1119. Google ScholarDigital Library
- Shi, W., Collins, E., and Karamcheti, V. 2003. Modeling object characteristics of dynamic web content. J. Para. Distrib. Comput. Google ScholarDigital Library
- Squillante, M., Woo, B., and Zhang, L. 2001. Analysis of queues under correlated arrivals with applications to Web server performance. Perf. Eval. Rev. 28, 4 (Mar.), 41--43. Google ScholarDigital Library
- Sriram, K. and Whitt, W. 1986. Characterizing superposition arrival processes in packet multiplexers for voice and data. IEEE J. Sel. Areas Comm. 4, 6 (Sept.), 833--846.Google ScholarDigital Library
- Starobinski, D. and Sidi, M. 2000. Stochastically bounded burstiness for communication networks. IEEE Trans. Info. Theory 46, 1 (Jan.), 206--212. Google ScholarDigital Library
- Tantawi, A. and Towsley, D. 1985. Optimal static load balancing in distributed computer systems. J. ACM 32, 2, 445--465. Google ScholarDigital Library
- Tantawi, A., Wolf, J., and Towsley, D. 1988. Optimal allocation of multiple class resources in computers systems. In Proceedings of the Sigmetrics, 253--260. Google ScholarDigital Library
- Urgaonkar, B., Pacifici, G., Shenoy, P., Spreitzer, M., and Tantawi, A. 2005. An analytical model for multi-tier internet services and its applications. In Proceedings of Sigmetrics 2005, 291--302. Google ScholarDigital Library
- Villela, D., Pradhan, P., and Rubenstein, D. 2004. Provisioning servers at the applicatior tier for e-commerce systems. In Proceedings of the International Workshop on Quality of Service (2004). Montreal, Canada.Google Scholar
- Wolf, J. and Yu, P. 2001. On balancing the load in a clustered web farm. ACM Trans. Internet Tech. 1, 2 (Nov.), 231--261. Google ScholarDigital Library
- Yaron, O. and Sidi, M. 1993. Performance and stability of communication networks via robust exponential bounds. IEEE/ACM Trans. Netw. 1, 3, 372--385. Google ScholarDigital Library
- Yashkov, S. 1983. A derivation of response time distribution for a M/G/1 processor sharing queue. Problems of Control and Information Theory 12, 133--148.Google Scholar
- Yashkov, S. F. 1987. Processor-sharing queues: Some progress in analysis. Queueing Syst. 2, 1--17. Google ScholarDigital Library
- Zhang, Z., Kurose, J., and Towsley, D. 1995. Statistical analysis of generalized processor sharing scheduling discipline. IEEE J. Sel. Areas in Comm. 13, 6 (Aug.), 1071--1080. Google ScholarDigital Library
- Zwart, B. and Boxma, O. 2000. Sojourn time asymptotics in the M/G/1 processor sharing queue. Queueing Syst. 35, 141--166. Google ScholarDigital Library
Index Terms
- Provisioning servers in the application tier for e-commerce systems
Recommendations
The internet vs e-commerce servers: when will server performance matter?
CASCON '98: Proceedings of the 1998 conference of the Centre for Advanced Studies on Collaborative researchThe cycle time of an Internet based online shopper includes time at an electronic commerce (e-commerce) server to gather information and purchase products, download time to transfer data over the Internet, and think time for interpreting the results of ...
A class-based scheme for E-commerce web servers: Formal specification and performance evaluation
This paper is an investigation into the performance of E-commerce applications. E-commerce has become one of the most popular applications of the web as a large population of web users is now benefiting from various on-line services including product ...
Analysis of Queueing Systems with Synchronous Single Vacation for Some Servers
We study a multi-server M/M/c type queue with a single vacation policy for some idle servers. In this queueing system, if at a service completion instant, any d (d ⩽c) servers become idle, these d servers will take one and only one vacation together. ...
Comments