James Griffioen and Randy Appleton. The design, implementation, and evaluation of a predictive caching file system. Technical Report CS--264-96. University of Kentucky, June 1996.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....that with no additional fetches, this policy works about as well as the best fixed policy, chosen a posteriori.Wedefinea refetch as a fetch of a previously seen object that is favored by the current policy but was discarded from the real cache at some time prior. Refetching is like prefetching [23, 8, 13], except that the refetched objects have recently been discarded. With refetching, the master policy can outperform the best fixed policy. In particular, when all required objects are refetched continuously, it has a 15 23 lower miss rate than the best fixed policy, and almost the same ....

....there need not be a single governing policy. Viewing the experts predictions as rankings makes them easy to combine. How rankings can be computed, combined, used to facilitate replacements, and suggest refetches, will be the topic of the following subsections. Refetches are like prefetches [23, 8, 13] except that the objects fetched have necessarily resided in the real cache at some time prior. More precisely, we define a refetch as a fetch of a previously seen object that was kept in the virtual cache(s) of higher weight expert policies, but was discarded from the real cache. Based on these ....

James Griffioen and Randy Appleton. The design, implementation, and evaluation of a predictive caching file system.

....on the object s URL as well as a digital signature of the object. Differencing was applied mainly to output of CGI scripts. In Prefetching, an object that might be needed in the future is fetched in advance. The utility of prefetching is studied in [13] using a statistical algorithm described in [7]. The bounds of latency reduction from caching and prefetching, based on search for objects with the same URL, is studied in [11] The authors found that caching and prefetching could reduce latency by at best 26 and 57 respectively. III. OUR CACHE BASED COMPACTION TECHNIQUE In the following, ....

James Griggioen and Randy Appleton. The design, implementation, and evaluation of a predictive caching file system. Technical Report CS-264-96, Department of Computer Science, University of Kentucky, Lexington, KY, June 1996.

....and indirectly, since the aging of files will affect the request frequencies of the approaches mentioned above. Padmanabhan and Mogul mention the need for (explicit) time modeling but show no solution for the problem [13] Some general considerations about document aging can also be found in [8]. The prediction scenario presented in this paper is based on our former approach described in [9] The main prediction ideas and algorithms are similar to the scheme of [10] We will strive here to refine the test scenario in order to present an adequate and explicit modeling of time and of the ....

James Griffioen, Randy Appleton, The design, Implementation, and Evaluation of a Predictive Caching File System, CS-Department University of Kentucky, CS-264-96, 1996.

....on the object s URL as well as a digital signature of the object. Differencing was applied mainly to output of CGI scripts. In Prefetching, an object that might be needed in the future is fetched in advance. The utility of prefetching is studied in [13] using a statistical algorithm described in [7]. The bounds of latency reduction from caching and prefetching, based on search for objects with the same URL, is studied in [11] The authors found that caching and prefetching could reduce latency by at best 26 and 57 respectively. III. OUR CACHE BASED COMPACTION TECHNIQUE In the following, ....

James Griggioen and Randy Appleton. The design, implementation, and evaluation of a predictive caching file system. Technical Report CS-264-96, Department of Computer Science, University of Kentucky, Lexington, KY, June 1996.

No context found.

James Griffioen and Randy Appleton. The design, implementation, and evaluation of a predictive caching file system. Technical Report CS--264-96. University of Kentucky, June 1996.

No context found.

James Griffioen and Randy Appleton. The design, implementation, and evaluation of a predictive caching file system. Technical Report CS--264-96. University of Kentucky, June 1996.

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