P. Krishnan and B. Sugla, "Utility of co-operating web proxy caches," in Proc. 7th Int. World Wide Web Conf., Brisbane, Australia, Apr. 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....congestion, proxy caching also alleviates server load and service latency. For even greater performance, these proxy caches can be organized into hierarchies, so that a local proxy can request a missing object from neighboring proxies instead of fetching it directly from the farther away server [3], 4] 5] Another way of using proxy caches is to place them in front of a particular server [6] These web proxies cache objects from usually one server to serve many clients at various locations in the network. The intension of this approach is to offload server load, instead of trying to ....

P. Krishnan and Binay Sugla, "Utility of co-operating web proxy caches," in Proc. of the 7th International WWW Conference, 1998.

....most frequently accessed documents thereby avoiding wide area network transfers and reducing client download latencies. To further improve document retrieval performance for web clients, several researchers have investigated the potential for cooperation among closely located web proxy caches [1, 2, 3, 4, 5, 6, 7, 8, 9]. However, a closer examination of these studies reveals that there is no clear consensus as to whether cooperation among proxy caches results in any additional bene t. Our goal in this work is to provide a closer examination at various forms of cooperation that have been studied in the literature ....

.... Cooperation amongst proxies is potentially bene cial in two ways: First, it reduces capacity misses by e ectively increasing the total cache size used to store documents [4, 7] Second, it reduces compulsory misses by opportunistically nding the missed object in another proxy s cache [5, 9]. While the former bene t is likely to be of little use if working set size is small, current trends in the Web suggest that workloads may potentially require large cache sizes. For example, the advent of content distribution networks that cache and distribute documents on behalf of heavily ....

[Article contains additional citation context not shown here]

P. Krishnan and Binay Sugla, \Utility of co-operating Web proxy caches," Computer Networks and ISDN Systems, vol. 30, no. 1-7, pp. 195-203, 1998.

....Moreover, the competition among electronic stores is motivating enhancements to existing services and the creation of new mechanisms to attract and keep customers, demanding even more computational power from servers. A common and successful strategy for improving WWW services is caching [17]. Frequently requested objects are stored in sites close to users, in terms of connectivity. Caching proxy servers reduce user perceived latency, network traffic, and load on the servers that have objects replicated. Although caching is a proven technology for static objects (i.e. HTML pages) ....

P.Krishnam and B. Sugla. Utility of co-operating web proxy caches. In Proceedings of WWW7, 1998.

....caching improves performance by sharing objects across multiple caches. This e ectively increases the cache store size. Additionally it increases the number of potential clients, which increases the probability that two clients will request the same object. An analysis of cooperative algorithms [14] concludes that the overhead in communication between cooperating caches may outweigh the bene t of additional hits provided by a neighboring cache. Dynamic policies provide better support in these circumstances. Cooperation should be used when the overhead is inconsequential compared to the ....

P. Krishnan and Binay Sugla. Utility of cooperating Web proxy caches. In ####### ######## ###### ##### #### ### ##########,volume 30(1{ 7) of ######## ######## ### #### #######, pages 195-203, Brisbane, Qld., Australia, 14-18 April 1998.

....CDNs. CDNs di er from hierarchical caches primarily in that they generalize the request routing function used to direct the miss stream from lower level (leaf) caches to the interior. Less signi cantly, CDNs may proactively populate interior caches, anticipating future requests. Previous research [3, 6, 7, 8, 9] has studied demandside bene ts to cooperative caching in the Web by analyzing traces collected in speci c caching environments. The goal of this paper is to adapt recent research results in demand side Web caching to study the interaction of supply side and demand side approaches to content ....

P. Krishnan and Binay Sugla. Utility of cooperating Web proxy caches. In Proceedings of the 7th International World Wide Web Conference, April 1997.

....is viable for a given set of parameters. Parameters for hit rates and cache speed ratio are independent of the cooperation mechanism, while and n depend upon the cache design. Table 5. 3 summarizes our estimates for , and n , where S is taken from reported local proxy hit rates [10] 21][41][2] 53] 68] and S and are taken from this work. For these values, the average speedup for a cooperative Web cache design with CHAPTER 5. VIABILITY ANALYSIS 83 Term Estimate Source 0.40 Approximate median of reported values 0.31 Geometric mean of real L3 hit rates from Table 5.1 ....

P. Krishnan and B. Sugla. Utility of co-operating Web proxy caches. Computer Networks and ISDN Systems, 30(1-7):105--203, Apr. 1998.

....caching improves performance by sharing objects across multiple caches. This e ectively increases the cache store size. Additionally it increases the number of potential clients, which increases the probability that two clients will request the same object. An analysis of cooperative algorithms [14] concludes that the overhead in communication between cooperating caches may outweigh the bene t of additional hits provided by a neighboring cache. Dynamic policies provide better support in these circumstances. Cooperation should be used when the overhead is inconsequential compared to the ....

P. Krishnan and Binay Sugla. Utility of cooperating Web proxy caches. In Seventh International World Wide Web Conference, volume 30(1{ 7) of Computer Networks and ISDN Systems, pages 195-203, Brisbane, Qld., Australia, 14-18 April 1998.

.... [FB96] GRC97] GCR97] GCR98] GC97] Gla94] GPB98] Gri97] GB97] GPV98] GNPV98] GS96a] GS96b] GS97] HMY97] Hed98] HN96] HWMS98] HSY98] HK97] HJWC98] IKY97] IST98] JC98] JDB96] JK97] JK98] Kah97] KKO98] WMS98a] KS98a] KW97a] KW97b] KW98] KMK99] KR99] KW99] KA99] KS98b] KLM97] KSW98b] KSW98a] LG98] LHC 98] LSCH98] LWS 99] LD99] LAJF98] Liu98] LC97] LC98] TB97] LOG96] LA94] Luo98] Mah99] MWE00] MEW00] MLB95] MR97] MS97] MSC98] Mar96] MC98] Mar99] MR98] WMFMA98] Mel96] MBV97] MA98] Mog95] Mog96] MDFK97a] MDFK97b] MJ98] ....

P. Krishnan and Binay Sugla. Utility of co-operating web proxy caches. In Proceedings of the Seventh International World Wide Web Conference, Brisbane, Australia, April 1998.

.... schemes like Harvest and Squid [9, 32] hash based schemes [21, 35] directory based schemes [16, 27, 33] and multicast based schemes [28, 34] Although each of these research efforts included a performance evaluation of the protocols proposed and a discussion of algorithm scalability, only [22] presents empirical evaluations of cooperation for small populations, and none present empirical or analytical evaluations of the effectiveness of their schemes for the large client populations found in a wide area setting. Using client traces, Krishnan et al. studied the utility of cooperation ....

....populations, and none present empirical or analytical evaluations of the effectiveness of their schemes for the large client populations found in a wide area setting. Using client traces, Krishnan et al. studied the utility of cooperation among three Bell Labs proxies with a small user population [22]. They concluded that cooperative Web caching can be useful, but that a cache manager was necessary to dynamically determine when to cooperate, because of the extra server load imposed by cooperation. This paper expands on these previous research efforts. We use trace based analysis to quantify ....

P. Krishnan and B. Sugla. Utility of co-operating Web proxy caches. In Proc. Seventh Int. World Wide Web Conf., April 1998.

....of documents, and reasons why documents are not cacheable. Finally, Section 7 summarizes our study and its results. 2 Previous Work Numerous recent studies of Web traffic have been performed. These studies include analyses of Web access traces from the perspective of browsers [11, 21] proxies [2, 4, 6, 10, 12, 15, 18, 19, 24], and servers [1, 3, 23] The earlier tracing studies were rather limited in request rate, number of requests, and diversity of population. The most recent tracing studies have been larger and generally more diverse. In addition to static analysis, some studies have also used trace driven cache ....

P. Krishnan and B. Sugla. Utility of co-operating web proxy caches. In Proc. Seventh InternationalWorld Wide Web Conference, April 1998.

....misses. Previous work has explored the performance of several removal policies: Least Recently Used (LRU) LeastFrequently Used (LFU) Perfect LFU [2] removal of large objects from caches [19] hybrid techniques utilizing estimation of client latencies [20] and cooperative caching techniques [9, 10]. These techniques demonstrate similar performance when applying the removal policy to all web objects. The algorithms each have special cases in which they perform well. An analysis of cooperating caches [10] concluded that dynamic policies that can turn on and o cooperation might be useful. ....

....utilizing estimation of client latencies [20] and cooperative caching techniques [9, 10] These techniques demonstrate similar performance when applying the removal policy to all web objects. The algorithms each have special cases in which they perform well. An analysis of cooperating caches [10] concluded that dynamic policies that can turn on and o cooperation might be useful. Ideally it would be useful to apply a removal policy to the set of objects for which it performs well. Caches determine whether documents are stale. Stale documents must be veri ed to determine that they are ....

Krishnan, P., and Sugla, B. Utility of cooperating Web proxy caches. In 7th International World Wide Web Conference (Brisbane, Qld., Australia, April 14-18 1998), vol. 30 of Computer Networks and ISDN Systems.

No context found.

P. Krishnan and B. Sugla, "Utility of co-operating web proxy caches," in Proc. 7th Int. World Wide Web Conf., Brisbane, Australia, Apr. 1997.

.... are at the edge of networks in the form of browser and proxy caches, the ends of high latency links, or as part of cache hierarchies [CDN 96, Squ] Signi cant research has gone into optimizing cache performance [CDN 96, WAS 96, NLA] co operation among several caches [CDN 96, KS97, MLB95, GRC97] and cache hierarchies [Squ, CDN 96, NL] Web servers are also replicated to achieve load balancing. An important method that is gaining momentum is caching inside the network [ZFJ97, DHS93, HM97] Danzig et al. DHS93] had observed the advantage of placing caches inside the ....

P. Krishnan and Binay Sugla. Utility of co-operating web proxy caches. In Proceedings of the 7th International World Wide Web Conference, Brisbane, Australia, April 1997.

.... the popular locations for caches are at the edge of networks in the form of browser and proxy caches, the ends of high latency links, or as part of cache hierarchies [8, 31] Significant research has gone into optimizing cache performance [8, 33, 29, 10] co operation among several caches [8, 23, 26, 17, 15], and cache hierarchies [31, 8, 28] Web servers are also replicated to achieve load balancing. Placing caches inside the network is becoming more popular [34, 35, 13, 20] Danzig et al. 13] had observed the advantage of placing caches inside the backbone rather than at its edges. They showed ....

P. Krishnan and B. Sugla. Utility of co-operating web proxy caches. In Proceedings of the 7th International World Wide Web Conference, Brisbane, Australia, Apr. 1997.

.... popular locations for caches are at the edge of networks in the form of browser and proxy caches, the ends of high latency links, or as part of cache hierarchies [8] 31] Significant research has gone into optimizing cache performance [8] 33] 29] 10] co operation among several caches [8] [23], 26] 17] 15] and cache hierarchies [31] 8] 28] Web servers are also replicated to achieve loadbalancing. Placing caches inside the network is becoming more popular [34] 35] 13] 20] Danzig et al. 13] had observed the advantage of placing caches inside the backbone rather than at ....

P. Krishnan and B. Sugla. Utility of co-operating web proxy caches. In Proceedings of the 7th International World Wide Web Conference, Brisbane, Australia, Apr. 1997.

.... are at the edge of networks in the form of browser and proxy caches, the ends of high latency links, or as part of cache hierarchies [CDN 96, Squ] Significant research has gone into optimizing cache performance [CDN 96, WAS 96, NLA] co operation among several caches [CDN 96, KS97, MLB95, GRC97] and cache hierarchies [Squ, CDN 96, NL] Web servers are also replicated to achieve loadbalancing. An important method that is gaining momentum is caching inside the network [ZFJ97, DHS93, HM97] Danzig et al. DHS93] had observed the advantage of placing caches inside the ....

P. Krishnan and Binay Sugla. Utility of co-operating web proxy caches. In Proceedings of the 7th International World Wide Web Conference, Brisbane, Australia, April 1997.

....form of browser and proxy caches, the end points of high latency links, or as part of cache hierarchies [5, 3] In cache hierarchies, the caches are configured to know their neighbors. Significant research has gone into optimizing cache performance [5, 21, 2, 6] co operation among several caches [5, 13, 17, 10, 9], and cache hierarchies [3, 5, 1] Web servers are also replicated to achieve load balancing. In this paper, we study the issues in using transparent en route caches (TERCs) in WANs. When using TERCs, only caches that are located along routes from a client to a server are used, transparently to ....

....in the last few years, making implementation feasible at backbone rates. Implementation at LAN speeds have been available for some time now. The emerging next generation of routers exhibit both high performance and rich functionality, such as support for virtual private networks and QoS [13, 14]. To achieve this, per flow queueing and fast IP filtering are incorporated into the router s hardware [14, 16] An example of products that provide such functionality today is Lucent s PacketStar 6400 IP switch 1 that incorporates the above features in hardware to support up to 128 T3 rate line ....

P. Krishnan and B. Sugla. Utility of co-operating web proxy caches. In Proceedings of the 7th International World Wide Web Conference, Brisbane, Australia, Apr. 1997.

No context found.

P. Krishnan and B. Sugla, "Utility of co-operating Web proxy caches," Computer Networks and ISDN Systems, vol. 30, 1998.

No context found.

P. Krishnan and B. Sugla. Utility of co-operating Web proxy caches. Computer Networks and ISDN Systems, 30(1-7):105--203, Apr. 1998.

No context found.

P. Krishnan and B. Sugla, "Utility of Cooperating Web Proxy Caches," in Proc. World-Wide Web Conf., pp. 195-203, April 1998.

No context found.

P. Krishnan and B. Sugla, \Utility of Cooperating Web Proxy Caches," in Proceedings World-Wide Web Conference, pp. 195-203, April 1998.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC