We propose a new design for highly concurrent Internet services, whichwe call the staged event-driven architecture (SEDA). SEDA is intended

The long-term success of the World Wide Web depends on fast response time. People use the Web to access information from remote sites, but do not like to wait long for their results. The latency of retrieving a Web document depends on several factors such as the network bandwidth, propagation time and the speed of the server and client computers. Although several proposals have been made for reducing this latency, it is difficult to push it to the point where it becomes insignificant. This motivates our work, where we investigate a scheme for reducing the latency perceived by users by predicting and prefetching files that are likely to be requested soon, while the user is browsing through the currently displayed page. In our scheme the server, which gets to see requests from several clients, makes predictions while individual clients initiate prefetching. We evaluate our scheme based on trace-driven simulations of prefetching over both high-bandwidth and low-bandwidth links.

The workload of the global Internet is dominated by the Hypertext Transfer Protocol (HTTP), an application protocol used by World Wide Web clients and servers. Simulation studies of this environment will require a model of the traffic patterns of the World Wide Web, in order to investigate the performance aspects of this increasingly popular application. We have developed an empirical model of network traffic produced by HTTP. Instead of relying on server or client logs, our approach is based on gathering packet traces of HTTP network conversations. Through traffic analysis, we have determined statistics and distributions for higher-level quantities such as the size of HTTP items retrieved, the number of items per "Web page", think time, and user browsing behavior. These quantities form a model can then be used by simulations to mimic World Wide Web network applications in wide-area IP internetworks. Keywords: World Wide Web, HTTP, traffic model, traffic measurements, workload, Interne...

Prefetching and caching are techniques commonly used in I/O systems to reduce latency. Many researchers have advocated the use of caching and prefetching to reduce latency in the Web. We derive several bounds on the performance improvements seen from these techniques, and then use traces of Web proxy activity taken at Digital Equipment Corporation to quantify these bounds. We found that for these traces, local proxy caching could reduce latency by at best 26%, prefetching could reduce latency by at best 57%, and a combined caching and prefetching proxy could provide at best a 60% latency reduction. Furthermore, we found that how far in advance a prefetching algorithm was able to prefetch an object was a significant factor in its ability to reduce latency. We note that the latency reduction from caching is significantly limited by the rapid changes of objects in the Web. We conclude that for the workload studied caching offers moderate assistance in reducing latency. Prefetching can of...

Web, workload characterization, performance, servers, caching, World Cup © Copyright Hewlett-Packard Company 1999 This paper presents a detailed workload characterization study of the 1998 World Cup Web site. Measurements from this site were collected over a three month period. During this time the site received 1.35 billion requests, making this the largest Web workload analyzed to date. By examining this extremely busy site and through comparison with existing characterization studies we are able to determine how Web server workloads are evolving. We find that improvements in the caching architecture of the World-Wide Web are changing the workloads of Web servers, but that major improvements to that architecture are still necessary. In particular, we uncover evidence that a better consistency mechanism is required for World-Wide Web caches.

In network (e.g., Web) servers, it is often desirable to isolate the performance of different classes of requests from each other. That is, one seeks to achieve that a certain minimal proportion of server resources are available for a class of requests, independent of the load imposed by other requests. Recent work demonstrates how to achieve this performance isolation in servers consisting of a single, centralized node; however, achieving performance isolation in a distributed, cluster based server remains a problem. This paper introduces a new abstraction, the cluster reserve, which represents a resource principal in a cluster based network server. We present a design and evaluate a prototype implementation that extends existing techniques for performance isolation on a single node server to cluster based servers. In our design, the dynamic cluster-wide resource management problem is formulated as a constrained optimization problem, with the resource allocations on individual machin...

We present a scalable architecture for content-aware request distribution in Web server clusters. In this architecture, a level-4 switch acts as the point of contact for the server on the Internet and distributes the incoming requests to a number of back-end nodes. The switch does not perform any content-based distribution. This function is performed by each of the back-end nodes, which may forward the incoming request to another back-end based on the requested content. In terms of scalability, this architecture compares favorably to existing approaches where a front-end node performs content-based distribution. In our architecture, the expensive operations of TCP connection establishment and hando are distributed among the back-ends, rather than being centralized in the front-end node. Only a minimal additional latency penalty is paid for much improved scalability. We have implemented this new architecture, and we demonstrate its superior scalability by comparing it to a system tha...

World-Wide Web server over the course a year and a half. This paper presents a longitudinal look at various network path properties, as well as the implementation status of various protocol options and mechanisms. In particular, this paper considers how WorldWide Web clients utilize TCP connections to transfer web data; the deployment of various TCP and HTTP options; the range of round-trip times observed in the network; packet sizes used for WWW transfers; the implications of the measured advertised window sizes; and the impact of using larger initial congestion window sizes. These properties/mechanisms and their implications are explored. An additional goal of this paper is to provide information to help researchers better simulate and emulate realistic networks.

An abridged version of this paper appeared in the Proceedings

Much work on the performance of Web proxy caching has focused on high-level metrics such as hit rates, but has ignored low-level details such as "cookies," aborted connections, and persistent connections between clients and proxies as well as between proxies and servers. These details have a strong impact on performance, particularly in heterogeneous bandwidth environments where network speeds between clients and proxies are significantly different than speeds between proxies and servers. We evaluate through detailed simulations the latency and bandwidth effects of Web proxy caching in such environments. We drive our simulations with packet traces from two scenarios: clients connected through slow dialup modems to a commercial ISP, and clients on a fast LAN in an industrial research lab. We present three main results. First, caching persistent connections at the proxy can improve latency much more than simply caching Web data. Second, aborted connections can waste more bandwidth than that saved by caching data. Third, cookies can dramatically reduce hit rates by making many documents effectively uncachable.