A. Dingle, T. Partl, "Web Cache Coherence", Proc. 5th Intl. WWW Conf., Paris, 1996.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....hits performed by these replacement strategies could be unacceptably high [6, 10] These studies propose different techniques to solve the problems of the stale hits. Two categories of methods have emerged as a potential solution for the problem of Web cache coherence the strong cache coherence [11] and the weak cache coherence [10] The cost and the efficiency of the two categories are still a controversial issue. The strong coherence strategies guarantee that no stale hits are performed by the cache server, while the weak strategies minimize the probability of these stale hits. The ....

....documents. The motivation behind this approach is the fact that cache servers have access to both the access history and the state of the cached documents; the cache can thus perform some statistics to predict network traffic. Methods such as push caching [12] or document refreshing (prefetching) [11] are simple examples of server initiated strategies. Using a simple simulation model and real workloads, we compare in this article five replacement strategies, namely, the LRU (the least recently used document is removed first) the LFU (the least frequently used document is removed first) the ....

[Article contains additional citation context not shown here]

Dingle, A., and T. Partl. 1996. Web cache coherence. Paper presented at the Proceedings of the Conference on Computer Networks and ISDN, pp. 907--20, May, Paris.

....of caching in the World Wide Web. These efforts fall into two categories: i) weak consistency mechanisms, where cached data can be outof sync with servers, and (ii) strong consistency mechanisms, where cached data is always up to date with servers. Examples of the former include client polling [13], where a client periodically polls the server for updates to the data, and adaptive time to live (TTL) values [32] where objects are refreshed based on the rate of change of the data. Examples of the latter include server invalidation [23] where the server sends sends invalidation messages to ....

A. Dingle and T. Partl. Web cache coherence. In Proc Fifth Intnl WWW Conference, May 1996.

....policies. We conclude in Section 7 with a summary and future research directions. 2. RELATED WORK Recent work addressed validation latency incurred on freshness misses, including transferring stale cached data from the cache to the client s browser while the data validity is being veri ed [14] or while the modi ed portion (the delta ) is being computed [15] These schemes, however, may require browser support and are e ective only when there is limited bandwidth between end user and cache (such as with modem users) Some or all freshness misses can be eliminated via stronger cache ....

A. Dingle and T. Partl, \Web cache coherence," in Proceedings of the Fifth International World Wide Web Conference, May 1996. http://www5conf.inria.fr/fich_html/papers/P2/Overview.html.

....(ttl) eld, either speci ed by the origin server, or computed by the web cache based on last modi cation time. The object is declared stale when its ttl expires. HTTP 1. 1 allows clients to control the expiration policy for requested objects, such as the degree of acceptable staleness [8]. For the purposes of this paper, we assume the existence of a binary valued freshness function that computes the freshness of an object based on factors such as time to live, time of last modi cation, time of object fetch from the origin server, and the current time. 2.2 Pastry A number of ....

A. Dingle and T. Partl. Web cache coherence. In Proceedings of the fth International World Wide Web Conference, Paris, May 1996.

....integrated in the replacement policy. The user information may concern (1) the subjective estimation of further reuse of a query result, 2) a preference for results coming from certain data repositories, 3) a fixed requirement of the staleness of the results returned by the cache(as claimed in [15]) Nevertheless, those preferences were never taken into account in caches so far, because the average user can not provide reasonable information, which will improve the percepted performance of the cache. On the other hand, it is difficult to integrate different preferences in a multi users ....

....we can suppose that our cached data is not incorrect but contains possibly not all answers currently available at the information source. In this situation, the problem of cache coherence is not a prevailing problem. Nevertheless, let s consider the coherency mechanisms used in web caches [6, 15]. In the first general approach the cache verifies for each access, whether the cached data is up todate. This mechanism has advantages in Web caches (page caches) as one can use the header If Modified Since: date in a HTTP GET message. The cost of this verification is less than retrieving once ....

A. Dingle, T.Partl. Web cache coherence. In Proc. Computer Networks and ISDN Systems, N. 28, pp.907 - 920, 1996.

....of our discussion, stale or defunct data refers to locally cached objects which are either (a) no longer are equivalent to the object on the originating server (a phenomenon discovered through comparison of object checksums) or (b) have since become obsolete. As identified by Dingle and Partl [10], there are four well known cache consistency maintenance techniques to deal with detecting such instances: client polling, invalidation callbacks, time to live, and ifmodified since. With client polling, caches timestamp each cached object and periodically compare the cached object timestamp ....

A. Dingle and T. Partl. Web cache coherence. Fifth International World Wide Web Conference, 1996.

....defunct data. For purposes of our discussion, stale or defunct data refers to locally cached objects which are either: No longer equivalent to the object on the originating server (a phenomenon discovered through comparison of object checksums) Obsolete As identified by Dingle and Partl [17], there are four well known cache consistency maintenance techniques to deal with detecting such instances: client polling, invalidation callbacks, time to live, and if modified since. With client polling, caches timestamp each cached object and periodically compare the cached object timestamp ....

A. Dingle and T. Partl, "Web Cache Coherence," 5th Int'l. World Wide Web Conf., 1996.

....policies. We conclude in Section VII with a summary and future research directions. II. RELATED WORK Recent work addressed validation latency incurred on freshness misses, including transferring stale cached data from the cache to the client s browser while the data validity is being verified [13] or while the modified portion (the delta ) is being computed [14] These schemes, however, may require browser support and are effective only when there is limited bandwidth between end user and cache (such as with modem users) Some or all freshness misses can be eliminated via stronger cache ....

A. Dingle and T. Partl, "Web cache coherence," in Proceedings of the Fifth International World Wide Web Conference, May 1996.

.... Proposed approaches to reduce user perceived latency incurred on validation requests included validating objects prior to predicted requests [18, 9] and server driven validations [20, 19] Other studies proposed transferring stale cached data to a client while the data validity is being verified [12] or while the modified portion (the delta ) is being computed [2] These approaches are of interest to us here since the issues of validation latency and DNS latency are conceptually related. Although considerable research targeted validation latency for content caching, it seems that no ....

A. Dingle and T. Partl. Web cache coherence. In Proceedings of the Fifth International World Wide Web Conference, May 1996.

....reducing Web access latency. The idea of content type specific compression is similar to [6] Caching is frequently used to improve the performance of distributed systems. Caching algorithms search for identical object. This topic has been studied extensively in the literature, see for example [4], 5] 8] and [14] Differencing compares an earlier version of an object to the current version. Usually, only two objects of the same URL or output of CGI script with different parameters are considered. Some of the differencing algorithm used are UNIX diff and vdelta [10] In [2] the issue ....

Adam Dingle and Tomas Partl. Web cache coherence. Fifth International World Wide Web Conference, May 1997.

....a list of documents to be validated whenever it forwards requests to servers. According to simulation results PCV reduces the proxy server communication while maintaining close tostrong cache coherency. Dingle and Partl propose a number of improvements to the TTL cache consistency mechanism in [40]. One of these is to base age calculation of age based TTLs not on the retrieval time stamp of the object but on a time stamp at which it was last known to be non stale. This differentiation becomes necessary in cache meshes where both time stamps are equal only at 11 the origin server of the ....

Adam Dingle and Tomas Partl. Web Cache Coherence. In Fifth International World-Wide Web Conference, Paris, France, May 6-10 1996. W3C.

No context found.

A. Dingle, T. Partl, "Web Cache Coherence", Proc. 5th Intl. WWW Conf., Paris, 1996.

No context found.

A. Dingle and T. Partl. Web cache coherence. In Proc Fifth International WWW Conference, May 1996, Paris, France.

No context found.

Adam Dingle and Toms Prtl. Web cache coherence. Proceedings of the Fifth International World Wide Web Conference on Computer Networks and ISDN Systems. May 1996.

No context found.

Adam Dingle and Toms Prtl. Web cache coherence. Proceedings of the Fifth International World Wide Web Conference on Computer Networks and ISDN Systems. May 1996.

No context found.

A. Dingle, T.Partl. Web cache coherence. In Proc. Computer Networks and ISDN Systems, N. 28, pp.907 - 920, 1996.

No context found.

A. Dingle and T. Partl. Web cache coherence. In Computer Networks & ISDN Systems, Vol.28, pp.907-920, May 1996.

No context found.

Dingle, A. and Partl, T., "Web cache coherence," Fifth International World Wide Web Conference, 1996.

No context found.

Adam Dingle and Thomas Partl. Web cache coherence. In Proceedings of the 5th International WWW Conference, May 1996.

No context found.

A. Dingle, T. Partl, "Web Cache Coherence", Proc. 5th Intl. WWW Conf., Paris, 1996.

No context found.

A. Dingle and T. Partl. Web cache coherence. In Proc Fifth Intnl WWW Conference, May 1996.

No context found.

A. Dingle and T. Partl. Web cache coherence. Proc. 5th WWW Conf., 1996.

No context found.

Adam Dingle. Web cache coherence. Proc. of the Fifth Intl. World Wide Web Conference, 1996.

No context found.

A. Dingle and T. Partl, "Web Cache Coherence," Computer Networks and ISDN Systems, 1996, Vol.28, No.7-11, pp.907-920.

No context found.

A. Dingle and T. Partl. Web cache coherence. In Proceedings of 5th WWW Conference (journal version: IJCN), 1997.

First 50 documents

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