The potential costs and benefits of long-term prefetching for content distribution

Abstract

This paper examines the costs and potential benefits of long-term prefetching for content distribution. In contrast with traditional short-term prefetching, in which caches use recent access history to predict and prefetch objects likely to be referenced in the near future, long-term prefetching uses long-term steadystate object access rates and update frequencies to identify objects to replicate to content distribution locations. Compared to demand caching, long-term prefetching increases network bandwidth and disk space costs but may benefit a system by improving hit rates. Using analytic models and trace-based simulations, we examine several algorithms for selecting objects for long-term prefetching. We find that although the web's Zipf-like object popularities makes it challenging to prefetch enough objects to significantly improve hit rates, systems can achieve significant benefits at modest costs by focusing their attention on long-lived objects.

Performance analysis of a client-side caching/prefetching system for Web traffic

“…Several studies have demonstrated the effectiveness of prefetching in addressing the limitations of passive caching (e.g., [14], [17], [22], [23], [27], [31], [32], [35], [42], [46], [49]). Prefetched documents may include hyperlinked documents that have not been requested yet as well as dynamic objects [37], [42].…”

Objective-greedy algorithms for long-term web prefetching

“…3. The paper also proves that the algorithms based on Good Fetch [14] and on the APL characteristic [9], although using different criteria, are equivalent in terms of their choice of objects selected for prefetching.…”

“…This criterion was first proposed by Jiang et al [9] as a performance metric in terms of both hit rate and bandwidth usage. Jiang et al [9] compared different algorithms, i.e., Prefetch by Popularity [12], Good Fetch [14], APL characteristic [9], and Lifetime [9], using this criterion, but did not give any algorithm to optimize this criterion. 2.…”

“…is the total number of web objects), which is the same as the cost incurred by the previous algorithms [14], [9], [12]. The performance comparison based on the simulations is summarized in Table 1 in Section 5.1.…”

“…Intuitively, to increase the hit rate, we want to prefetch those objects that are accessed most frequently; to minimize the bandwidth consumption, we may want to choose those objects with longer update intervals. This paper first reviews the popular prefetching algorithms based on Popularity [12], Good Fetch [14], APL characteristic [9], and Lifetime [9]. Their performance can be measured using the different criteria discussed.…”