Abstract
This paper analyzes quatitatively the impact of the load balance scheduling algorithms and the locality scheduling algorithms on the performance of Web cluster servers, and brings forward the Adaptive_LARD algorithm. Compared with the representative LARD algorithm, the advantages of the Adaptive_LARD are that: (1) it adjusts load distribution among the back-ends through the idea of load balancing to avoid learning steps in the LARD algorithm and reinforce its adaptability; (2) by distinguishing between TCP connections accessing disks and those accessing cache memory, it can estimate the impact of different connections on the back-ends’ load more precisely. Performance evaluations suggest that the proposed method outperforms the LARD algorithm by up to 14.7%.
Similar content being viewed by others
References
Fox A, Gribble S D, Chawathe Yet al. Cluster-based scalable network services. InProc. the 16th Symposium on Operating Systems Principles, Saint Malo, France, 1997, pp.78–91.
Cardellini V, Colajanni M, Yu P S. Dynamic load balancing on Web-server systems.IEEE Internet Computing, May–June 1999, 3(3): 28–39.
Cohen A, Rangarajan S, Slye H. On the performance of TCP splicing for URL-aware redirection. InProc. the 2nd Usenix Symposium on Internet Technologies and Systems, Boulder, CO, Oct. 1999, pp.117–125.
Oliver Spatschecket al. Optimizing TCP forwarder performance. InIEEE/ACM Trans. Networking, April 2000, 8(2): 146–157.
Yang C S, Luo M Y. Efficient support for contentbased routing in Web server clusters. InProc. the 2nd Usenix Symposium on Internet Technologies and Systems, Boulder, CO, Oct. 1999.
Cisco Systems Inc. LocalDirector. http://www.cisco.com.
Guerney D H Hunt, Germán S Goldszmidt, Richard P Kinget al. Network dispatcher: A connection router for scalable Internet services.Computer Networks and ISDN Systems, Apr. 1998, 30(1–7): 347–357.
Pai V S, Aron M, Banga Get al. Locality-aware request distribution in cluster-based network servers. InProc. the 8th Conference on Architectural Support for Programming Languages and Operating Systems, San Jose, CA, Oct. 1998, pp.205–216.
Lei Yingchun, Zhang Song, Li Guojie. Segregation scheduling policy in Web cluster servers.Journal of Computer Research and Development, Sept. 2002, 39(9): 1093–1098.
Zhu H. Scheduling optimization for resource-intensive Web requests on server clusters. InProc. 11th Annual ACM Symposium on Parallel Algorithms and Architectures (SPAA’99), June 1999, pp.13–22.
Lei Yingchun, Li Guojie. High efficient L5-Dispatcher based on requests’ contents.Journal of Computer Research and Development, Feb. 2002, 39(2): 183–191.
Lei Yingchun, Zhang Song, Li Guojie. The distributed L5-Dispatchers.Journal of Computer Research and Development, Mar 2003, 40(3): 1153–1161.
Bianchini R, Carrera E V. Analytical and experimental evaluation of cluster-based network servers.World Wide Web Journal, Dec. 2000, 3(4): 215–229.
Bunt R B, Eager D L, Oster G M, Williamson C L. Achieving load balance and effective caching in clustered Web servers. InProc. 4th International Web Caching Workshop, San Diego, CA, March 1999.
Arlitt M, Williamson C L. Web server workload characterization: The search for invariants. InProc. the ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems, May 1996.
Barford P, Crovella M. Generating representative Web workload for network and server performance evaluation. Technical Report 1997-006, Computer Science Department, Boston University, 1997.
Lei Yingchun, Zhang Song, Li Guojie. Analyzing the relationship between scheduling algorithm and the performance of Web cluster servers.Journal of Computer Research and Development, Mar. 2003, 40(3): 483–492.
Lei Yingchun, Li Guojie. A high performance Web server based on the ASPED model.Journal of Computer Research and Development, July 2001, 38(7): 769–777.
Aron M, Druschel P, Zwaenepoel W. Efficient support for P-HTTP in cluster-based Web servers. InProc. the USENIX 1999 Annual Technical Conference, Monterey, CA, June 1999, pp.185–198.
Aron M, Sanders D, Druschel P, Zwaenepoel W. Scalable content-aware request distribution in cluster-based network servers. InProc. the USENIX 2000 Annual Technical Conference, San Diego, CA, June 2000.
Genova Z, Christensen K. Challenges in URL switching for implementing globally distributed Web sites. InProc. the Workshop on Scalable Web Services, August 2000, pp.89–94.
Web Characterization Repository. http://researchsmp2. cc.vt.edu/
Menasce D A, Almeida V F. Capacity Planning for Web Performance, Prentice-Hall, 1998.
IBM Deskstar 60GXP. http://www.storage.ibm.com/2001.
Standard performance evaluation Corp. (SPEC), SPEC-Web99 Benchmark. http://www.specbench.org/osg/Web99, 1999.
Martin A, Tai J Workload characterization of the 1998 World Cup Web Site. http://www.hpl hp. com/techreports/1999/HPL-1999-35R1.html.
Author information
Authors and Affiliations
Corresponding author
Additional information
This work was supported by the National High-Tech 863 Program of China (Grant No.2002AA1Z2102).
Lei YingChun received the B.S. degree and the M.S. degree, both in computer science, from Wuhan University in 1995 and 1998 respectively, and the Ph.D. degree in computer system architecture in 2002 from Institute of Computing Technology, the Chinese Academy of Sciences. He is currently with the National Research Center for High Performance Computers at Institute of Computing Technology, the Chinese Academy of Sciences. His research interests include operating system, networking server, streaming server, multimedia communication, and grid.
Gong YiLi received her B.S. degree in computer science from Wuhan University in 1998 and began to pursue the M.S. and the Ph.D. degrees at Institute of Computing Technology, the Chinese Academy of Sciences in 2000. Her main research interests include high performance servers, resource representation, scheduling and quality of service in grid and peer-to-peer.
Zhang Song received the B.S. degree and the M.S. degree, both in computer science, from the Xian Jiaotong University in 1992 and 1995 respectively, and the Ph.D. degree in computer system architecture in 1999 from Institute of Computing Technology, the Chinese Academy of Sciences. He is currently with the National Research Center for High Performance Computers at Institute of Computing Technology, the Chinese Academy of Sciences. His research interests include high-speed network, appliance cluster, streaming and multimedia communication.
Li GuoJie received his B.S. degree in physics from Peking University in 1968, M.S. degree in computer science from University of Science & Technology of China in 1981, and Ph.D. degree in E.E. from Purdue University, USA in 1985. He is a member of Chinese Academy of Engineering and the director of Institute of Computing Technology, the Chinese Academy of Sciences. His current research interests include high performance computer architecture, and MPP systems for bio-information processing.
Rights and permissions
About this article
Cite this article
Lei, Y., Gong, Y., Zhang, S. et al. Research on scheduling algorithms in Web cluster servers. J. Comput. Sci. & Technol. 18, 703–716 (2003). https://doi.org/10.1007/BF02945458
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02945458