In the span of only a few years, the Internet has experienced an astronomical increase in the use of specialized content delivery systems, such as content delivery networks and peer-to-peer file sharing systems. Therefore, an understanding of content delivery on the Internet now requires a detailed understanding of how these systems are used in practice. This paper examines content delivery from the point of view of four content delivery systems: HTTP web traffic, the Akamai content delivery network, and Kazaa and Gnutella peer-to-peer file sharing traffic. We collected a trace of all incoming and outgoing network traffic at the University of Washington, a large university with over 60,000 students, faculty, and staff. From this trace, we isolated and characterized traffic belonging to each of these four delivery classes. Our results (1) quantify the rapidly increasing importance of new content delivery systems, particularly peerto-peer networks, (2) characterize the behavior of these systems from the perspectives of clients, objects, and servers, and (3) derive implications for caching in these systems. 1

There is an increasing need to quickly and efficiently learn network distances, in terms of metrics such as latency or bandwidth, between Internet hosts. For example, Internet content providers often place data and server mirrors throughout the Internet to improve access latency for clients, and it is necessary to direct clients to the closest mirrors based on some distance metric in order to realize the benefit of mirrors. We suggest a scalable Internet-wide architecture, called IDMaps, which measures and disseminates distance information on the global Internet. Higher-level services can collect such distance information to build a virtual distance map of the Internet and estimate the distance between any pair of IP addresses. We present our solutions to the measurement server placement and distance map construction problems in IDMaps. We show that IDMaps can indeed provide useful distance estimations to applications such as closest-mirror selection. 1 Keywords: network service, distributed algorithms, scalability, modeling. 1

Content distribution networks (CDNs) are a mechanism to deliver content to end users on behalf of origin Web sites. Content distribution offloads work from origin servers by serving some or all of the contents of Web pages. We found an order of magnitude increase in the number and percentage of popular origin sites using CDNs between

Abstract—This paper presents a detailed analysis of traces of domain name system (DNS) and associated TCP traffic collected

Abstract not found

The Domain Name System (DNS) is a ubiquitous part of everyday computing, translating human-friendly machine names to numeric IP addresses. Most DNS research has focused on server-side infrastructure, with the assumption that the aggressive caching and redundancy on the client side are sufficient. However, through systematic monitoring, we find that client-side DNS failures are widespread and frequent, degrading DNS performance and reliability. We introduce CoDNS, a lightweight, cooperative DNS lookup service that can be independently and incrementally deployed to augment existing nameservers. It uses a locality and proximity-aware design to distribute DNS requests, and achieves low-latency, low-overhead name resolution, even in the presence of local DNS nameserver delay/failure. Using live traffic, we show that CoDNS reduces average lookup latency by 27-82%, greatly reduces slow lookups, and improves DNS availability by an additional ’9’. We also show that a widely-deployed service using CoDNS gains increased capacity, higher reliability, and faster start times. 1

Replication is a well-known technique to improve the accessibility of Web sites. It generally offers reduced client latencies and increases a site’s availability. However, applying replication techniques is not trivial, and various Content Delivery Networks (CDNs) have been created to facilitate replication for digital content providers. The

This thesis presents a new design and implementation of the DHash distributed hash table based on erasure encoding. This design is both more robust and more efficient than the previous replication-based implementation [15]. DHash uses

Multihoming is increasingly being employed by large enterprises and data centers as a mechanism to extract good performance from their provider connections. Today, multihomed end-networks can employ a variety of commercial route control products to optimize performance over multiple ISP links. However, little is known about the mechanisms employed by such products and their relative trade-offs. In this paper, we propose and evaluate a wide range practical schemes that could go into the design of a route control device and analyze their trade-offs. We implement the proposed schemes on a Linux-based Web proxy and perform a trace-based emulation of their relative performance benefits. We show that both passive and active monitoring based techniques are equally effective and could improve Web performance by about 25 % when compared to using a single provider. Another key observation is that the conventional practice of employing historical measurement samples to monitor and predict ISP performance could, in fact, result in sub-optimal performance. 1

Abstract—Replication of information among multiple servers is necessary to support high request rates to popular Web sites. We consider systems that maintain one interface to the users, even if they consist of multiple nodes with visible IP addresses that are distributed among different networks. In these systems, the first-level dispatching is achieved through the Domain Name System (DNS) during the address lookup phase. Distributed Web systems can use some request redirection mechanism as a second-level dispatching because the DNS routing scheme has limited control on offered load. Redirection is always executed by the servers, but there are many alternatives that are worth of investigation. In this paper, we explore the combination of DNS dispatching with redirection schemes that use centralized or distributed control on the basis of global or local state information. In the fully distributed schemes, DNS dispatching is carried out by simple algorithms because load sharing are taken by some redirection mechanisms that each server activates autonomously. On the other hand, in fully centralized schemes, redirection is used as a tool to enforce the decisions taken by the same centralized entity that provides the first-level dispatching. We also investigate some hybrid strategies. We conclude that the distributed algorithms are preferable over the centralized counterpart because they provide stable performance, can take content-aware dispatching decisions, can limit the percentage of redirected requests and, last, but not least, their implementation is much simpler than that required by the centralized schemes.