Forwarding Requests among Reverse Proxies
نویسندگان
چکیده
Reverse proxy caching is a technology deployed by many ISPs at the border routers of their backbones to improve performance of their Web hosting services. Currently, cooperation among reverse proxies (if any) is limited to sharing each other’s cache copies. In this paper, we propose to extend the cooperation by forwarding requests among cooperating reverse proxies. Instead of fetching objects from remote proxies, a proxy in this mechanism forwards requests to other proxies and tells them to send the objects directly to clients. The resulting “circular communication” (client→proxy→remote proxy→client) can be implemented in practice with a TCP hand-off among proxies. Request forwarding has serveral potential benefits: first, it can get the heavy traffic off an ISP’s own backbone quickly and make more backbone bandwidth available to accommodate more customers; second, it can offload busy proxies and achieve some load balancing; finally, by observing network delay of their previous interactions with clients, reverse proxies can use request forwarding to improve client-proxy network proximity and avoid congested networks. Using trace-driven simulations, we evaluated the first benefit by studying two policies. Preliminary results show a 13-35% backbone bandwidth reduction and the benefit of maintaining a dedicated output link at the content server.
منابع مشابه
SSLSARD: A Request Distribution Technique for Distributed SSL Reverse Proxies
—Although Secure Sockets Layer (SSL) and its successor Transport Layer Security (TLS) are the for transport layer security, their cryptographic operations tend to be highly CPU intensive. Web systems that support SSL/TLS often deploy several locally or globally distributed SSL reverse proxies in front of Web servers to offload SSL/TLS operations from Web servers and improve the execution perfo...
متن کاملبهینهسازی اجرا و پاسخ صفحات وب در فضای ابری با روشهای پیشپردازش، مطالعه موردی سامانههای وارنیش و انجینکس
The response speed of Web pages is one of the necessities of information technology. In recent years, renowned companies such as Google and computer scientists focused on speeding up the web. Achievements such as Google Pagespeed, Nginx and varnish are the result of these researches. In Customer to Customer(C2C) business systems, such as chat systems, and in Business to Customer(B2C) systems, s...
متن کاملIdentifying a framework for addressing the challenges inherent in the complexity of contemporary Web architecture . Locality in a Web of Streams
T he dramatic growth of the Web is well known and often discussed within the computing community. Furthermore, beyond size (measured in terms of content, users, or number of servers), the Web has also grown considerably in complexity since its inception. Broadly speaking, the Web has evolved from a simple collection of clients and servers to a much more complex arrangement, including new kinds ...
متن کاملLocation Proxies and Intermediate Node Forwarding for Practical Geographic Forwarding
Two main problems prevent the deployment of geographic forwarding in real systems: geographic forwarding requires that all nodes know their locations, and it has trouble routing around local dead ends. This paper presents practical solutions to each problem. The location proxy technique allows a node that does not know its location to find a nearby location aware node to use as a proxy for geog...
متن کاملImproving Learning-Based Request Forwarding in Resource Discovery through Load-Awareness
Request forwarding is an efficient approach in discovering resources in distributed systems because it achieves one of the main goals of distributed systems namely the scalability goal. Despite achieving reasonable scalability, this approach suffers from long response times to resource requests. Several solutions such as learning-based request forwarding have tried to improve the response time ...
متن کامل