Handling Multiple Bottlenecks in Web Servers Using Adaptive Inbound Controls

  • Thiemo Voigt
  • Per Gunningberg
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2334)


Web servers become overloaded when one or several server resources are overutilized. In this paper we present an adaptive architecture that prevents resource overutilization in web servers by performing admission control based on application-level information found in HTTP headers and knowledge about resource consumption of requests. In addition, we use an efficient early discard mechanism that consumes only a small amount of resources when rejecting requests. This mechanism first comes into play when the request rate is very high in order to avoid making uninformed request rejections that might abort ongoing sessions. We present our dual admission control architecture and various experiments that show that it can sustain high throughput and low response times even during high load.


Queue Length Admission Control Request Rate Token Bucket Bucket Size 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Abdelzaher T., Bhatti N.: Web Server QoS Management by Adaptive Content Delivery. Int. Workshop on Quality of Service (1999)Google Scholar
  2. 2.
    Abdelzaher T., Lu C.: Modeling and Performance Control of Internet Servers. Invited paper, IEEE Conference on Decision and Control (2000)Google Scholar
  3. 3.
    Almeida J., Dabu M., Manikutty A., Cao P.: Providing Differentiated Levels of Service in Web Content Hosting. Internet Server Performance Workshop (1999)Google Scholar
  4. 4.
    Arlitt M., Williamson C.: Web Server Workload Characterization: The Search for Invariants. Proc. of ACM Sigmetrics (1996)Google Scholar
  5. 5.
    Banga G., Druschel P.: Measuring the Capacity of a Web Server. USENIX Symposium on Internet Technologies and Systems (1997)Google Scholar
  6. 6.
    Barford P., Crovella M.: Generating Representative Web Workloads for Network and Server Performance Evaluation. Proc. of SIGMETRICS (1998)Google Scholar
  7. 7.
    Bhatti N., Friedrich R.: Web Server Support for Tiered Services. IEEE Network (1999) 36–43Google Scholar
  8. 8.
    Casalicchio E., Colajanni M.: A Client-Aware Dispatching Algorithm for Web Clusters Providing Multiple Services. 10th Int’l World Wide Web Conference (2001)Google Scholar
  9. 9.
    Cherkasova L., Phaal P.: Session Based Admission Control: A Mechanism for Improving the Performance of an Overloaded Web Server. Tech Report: HPL-98-119 (1998)Google Scholar
  10. 10.
    Eggert L., Heidemann J.: Application-Level Differentiated Services for Web Servers. World Wide Web Journal (1999) 133–142Google Scholar
  11. 11.
    Parekh S. et al.: Using Control Theory to Achieve Service Level Objectives in Performance Management. Int. Symposium on Integrated Network Management (2001)Google Scholar
  12. 12.
    Glad T., Ljung L.: Reglerteknik: Grundläggande teori (in Swedish). Studentlitteratur (1989)Google Scholar
  13. 13.
    Iyer R., Tewari V., Kant K.: Overload Control Mechansims for Web Servers. Performance and QoS of Next Generation Networks (2000)Google Scholar
  14. 14.
    Lu C. et al.: A Feedback Control Approach for Guaranteeing Relative Delays in Web Servers. Real-Time Technology and Application Symposium (2001)Google Scholar
  15. 15.
    Mindcraft: Webstone,
  16. 16.
    Alteon: Alteon Web OS Traffic Control Software.
  17. 17.
    Inktomi: Inktomi Traffic Server C-Class.
  18. 18.
    Cisco: Cisco LocalDirector.
  19. 19.
    Jamjoom H., Reumann J.: QGuard: Protecting Internet Servers from Overload. University of Michigan CSE-TR-427-00 (2000)Google Scholar
  20. 20.
    Bhoj P., Ramanathan S., Singhal S.: Web2K: Bringing QoS to Web Servers. Tech Report: HPL-2000-61 (2000)Google Scholar
  21. 21.
    Challenger J., Dantzig P., Iyengar A.: A Scalable and Highly Available System for Serving Dynamic Data at Frequently Accessed Web Sites. Proc. of ACM/IEEE SC 98 (1998)Google Scholar
  22. 22.
    Manley S., Seltzer M.: Web Facts and Fantasy. USENIX Symposium on Internet Technologies and Systems (1997)Google Scholar
  23. 23.
    Harchol-Balter M., Schroeder B., Agrawal M., Bansal N.: Size-based Scheduling to Improve Web Performance. (2002)
  24. 24.
    van de Ven A.: KHTTPd,
  25. 25.
    Aron M., Sanders D., Druschel P., Zwaenepoel W.: Scalable Content-aware Request Distribution in Cluster-based Network Servers. Usenix Annual Technical Conference (2000)Google Scholar
  26. 26.
    Pai V., Aron M., Banga G., Svendsen M., Druschel P., Zwaenepoel W., Nahum E.: Locality-aware Request Distribution in Cluster-based Network Servers. International Conference on Architectural Support for Programming Languages and Operating Systems (1998)Google Scholar
  27. 27.
    Zhang X., Barrientos M., Chen J., Seltzer M.: HACC: An Architecture for Cluster-Based Web Servers. Third Usenix Windows NT Symposium (1999)Google Scholar
  28. 28.
    Aron M., Druschel P., Zwaenepoel W.: Cluster Reserves: a Mechanism for Resource Management in Cluster-based Network Servers. Proc. of ACM SIGMET-RICS (2000)Google Scholar
  29. 29.
    Cardellini V., Calajanni M., Yu P.: Dynamic Load Balancing on Web-server Systems. IEEE Internet Computing (1999) 28–39Google Scholar
  30. 30.
    Wang Z.: Cachemesh: A Distributed Cache System for the World Wide Web. 2nd NLANR Web Caching Workshop (1997)Google Scholar
  31. 31.
    Voigt T., Tewari R., Freimuth D., Mehra A.: Kernel Mechanisms for Service Differentiation in Overloaded Web Servers. Usenix Annual Technical Conference (2001)Google Scholar
  32. 32.
    Voigt T., Gunningberg P.: Kernel-based Control of Persistent Web Server Connections, ACM Performance Evaluation Review (2001) 20–25Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Thiemo Voigt
    • 1
  • Per Gunningberg
    • 2
  1. 1.SICSKistaSweden
  2. 2.Uppsala UniversityUppsalaSweden

Personalised recommendations