Instantaneous Offloading of Transient Web Server Load 1

A modern web-hosting site is designed to handle load that is sometimes an order of magnitude greater than the average load. Such a site can be expensive and is underutilized most of the time. We describe a design and performance study of the web booster architecture, which reduces web server load during peak periods. A web booster, inserted between client and server, instantaneously decreases server processing costs for requests to static documents while keeping the processing on the origin server. Fast reaction to the load change and the fact that the booster does not hold resources during inactive periods allows a web booster to be time-shared among multiple web domains that physically reside on different servers. A web accelerator module on a web server interacts with the web booster. It uses network packet caching, optimizing the TCP protocol and offloading of computation to increase the rate at which a server can process client requests. This paper describes the web booster architecture and our implementation of the accelerator, which decreases the web server load by more than a factor of three.

[1]  W. Richard Stevens.  TCP/IP Illustrated, Volume 3: TCP for Transactions, HTTP, NNTP, and the UNIX Domain Protocolls , 1996 .

[2]  Willy Zwaenepoel,et al.  Flash: An efficient and portable Web server , 1999, USENIX Annual Technical Conference, General Track.

[3]  W. Richard Stevens.  TCP for transactions, HTTP, NNTP, and the UNIX domain protocols , 1996 .

[4]  Margo Seltzer,et al.  HACC: an architecture for cluster-based web servers , 1999 .

[5]  Willy Zwaenepoel,et al.  Efficient Support for P-HTTP in Cluster-Based Web Servers , 1999, USENIX Annual Technical Conference, General Track.

[6]  Douglas C. Schmidt,et al.  Techniques for developing and measuring high performance Web servers over high speed networks , 1998, Proceedings. IEEE INFOCOM '98, the Conference on Computer Communications. Seventeenth Annual Joint Conference of the IEEE Computer and Communications Societies. Gateway to the 21st Century (Cat. No.98.

[7]  Azer Bestavros,et al.  Distributed packet rewriting and its application to scalable server architectures , 1998, Proceedings Sixth International Conference on Network Protocols (Cat. No.98TB100256).

[8]  G. Voelker,et al.  On the scale and performance of cooperative Web proxy caching , 2000, OPSR.

[9]  Jeffrey C. Mogul,et al.  Scalable Kernel Performance for Internet Servers Under Realistic Loads , 1998, USENIX Annual Technical Conference.

[10]  Douglas C. Schmidt,et al.  Techniques for Developing and Measuring High-Performance Web Servers over ATM Networks , 1998 .

[11]  Carlos Maltzahn,et al.  Performance issues of enterprise level web proxies , 1997, SIGMETRICS '97.

[12]  Michael Dahlin,et al.  Design considerations for distributed caching on the Internet , 1999, Proceedings. 19th IEEE International Conference on Distributed Computing Systems (Cat. No.99CB37003).

[13]  Eric Levy-Abegnoli,et al.  Design and performance of a Web server accelerator , 1999, IEEE INFOCOM '99. Conference on Computer Communications. Proceedings. Eighteenth Annual Joint Conference of the IEEE Computer and Communications Societies. The Future is Now (Cat. No.99CH36320).

[14]  Sampath Rangarajan,et al.  On the Performance of TCP Splicing for URL-Aware Redirection , 1999, USENIX Symposium on Internet Technologies and Systems.

[15]  Seth Copen Goldstein,et al.  Active messages: a mechanism for integrating communication and computation , 1998, ISCA '98.

[16]  Alec Wolman,et al.  On the scale and performance of cooperative Web proxy caching , 1999, SOSP.

[17]  Daniel M. Dias,et al.  A scalable and highly available web server , 1996, COMPCON '96. Technologies for the Information Superhighway Digest of Papers.

[18]  David R. Karger,et al.  Web Caching with Consistent Hashing , 1999, Comput. Networks.

[19]  Li Fan,et al.  Web caching and Zipf-like distributions: evidence and implications , 1999, IEEE INFOCOM '99. Conference on Computer Communications. Proceedings. Eighteenth Annual Joint Conference of the IEEE Computer and Communications Societies. The Future is Now (Cat. No.99CH36320).

[20]  Thomas P. Brisco.  DNS Support for Load Balancing , 1995, RFC.

[21]  Paul Barford,et al.  Generating representative Web workloads for network and server performance evaluation , 1998, SIGMETRICS '98/PERFORMANCE '98.

[22]  Willy Zwaenepoel,et al.  IO-Lite: a unified I/O buffering and caching system , 1999, TOCS.

[23]  Seth Copen Goldstein,et al.  Active messages: a mechanism for integrating communication and computation , 1998, ISCA '98.

[24]  Mon-Yen Luo,et al.  Efficient Support for Content-based Routing in Web Server Clusters , 1999, USENIX Symposium on Internet Technologies and Systems.

[25]  Thorsten von Eicken,et al.  U-Net: a user-level network interface for parallel and distributed computing , 1995, SOSP.

[26]  Virgílio A. F. Almeida,et al.  Characterizing reference locality in the WWW , 1996, Fourth International Conference on Parallel and Distributed Information Systems.

[27]  Peter B. Danzig,et al.  A Hierarchical Internet Object Cache , 1996, USENIX ATC.

[28]  Peter Druschel,et al.  Better operating system features for faster network servers , 1998, PERV.

[29]  Willy Zwaenepoel,et al.  Scalable Content-aware Request Distribution in Cluster-based Network Servers , 2000, USENIX ATC, General Track.

[30]  Erich M. Nahum,et al.  Locality-aware request distribution in cluster-based network servers , 1998, ASPLOS VIII.

[31]  Anja Feldmann,et al.  Rate of Change and other Metrics: a Live Study of the World Wide Web , 1997, USENIX Symposium on Internet Technologies and Systems.

[32]  Anja Feldmann,et al.  Web proxy caching: the devil is in the details , 1998, PERV.

[33]  Trent Jaeger,et al.  High-Performance Caching With The Lava Hit-Server , 1998, USENIX ATC.