K. Grimsrud, J. Archibald, and B. Nelson, Multiple prefetch adaptive disk caching, IEEE Transactions on Knowledge and Data Engineering, February 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....is far more expensive than the one associated with deterministic techniques. We define our own statistical prediction technique in Chapter 5. Table method. To keep track of the inter dependency of block accesses a table was used to register blocks which will be accessed after a current block [Grimsrud et al. 1993]. Each referenced block has an associated weight which gives information about the probability of accessing the next block. Two other studies use the same algorithm [Shah and Kumar, 1995; Chee et al. 1997] In the area of CPU prefetching a so called reference prediction table was introduced to ....

....Strategies Although it would make sense to treat prefetched pages and demand pages differently few researchers have made this distinction. Both types of requests shared the same buffer pool with the LRU replacement strategy [Palmer and Zdonik, 1990; Fu and Patel, 1991; Palmer and Zdonik, 1991; Grimsrud et al. 1993; Pai and Varman, 1992; Curewitz et al. 1993] Independent of the type of request, all new pages get the highest priority. The same is true for pages that are predicted but already resident [Horspool and Huberman, 1987; Wilson et al. 1994] Chang and Katz, 1989; Chang, 1989] used two priority ....

[Article contains additional citation context not shown here]

Grimsrud, K., Archibald, J., and Nelson, B. (1993). Multiple Prefetch Adaptive Disk Caching. IEEE Knowledge and Data Engineering, 5(1):88--103.

....is the time factor. Smith [16] proposed two policies: a) prefetch only when a buffer fault occurred (demand prefetch) b) prefetch at any time (prefetch always) Prediction information must be efficiently stored on disk and accessed in memory. This information can be stored in tables ( 7] [8], 11] or Markov chains ( 5] 13] 3 The Prefetching Design 3.1 Prefetch Object Table OODBMSs can store and retrieve large, complex data structures which are nested and heavily interrelated. Examples of OODBMS applications are CASE, CAD, CAM and Office automation. These applications consists ....

K. S. Grimsrud, J. K. Archibald, and B. E. Nelson. Multiple Prefetch Adaptive Disk Caching. IEEE Knowledge and Data Engineering, 5(1), February 1993.

....were obtained from past accesses. In [7] a probability graph was used to predict the future file access, i.e. the probability that one file will be opened after another. In the WWW Bestavros [1] speculated about document accesses based upon their document interdependency probabilities. Grimsrud [8] keeps a record of the disk clusters which were accessed immediately after another cluster with an associated weight for the access probability. An interesting study to preload documents from tertiary storage was made by Kraiss [12] This work is probably most related to our work because it uses a ....

K.S. Grimsrud, J.K. Archibald, and B.E. Nelson. Multiple Prefetch Adaptive Disk Caching. IEEE Knowledge and Data Engineering, 5(1):88--103, February 1993.

....past accesses. In [7] a probability graph was used to predict the future file access, i.e. the probability that one file will be opened after another. In the World Wide Web [1] Bestavros speculated about document accesses based upon their document interdependency probabilities. Grimsrud et al. [8] keep a record of the disk clusters which were accessed immediately after another cluster with an associated weight for the access probability. An interesting study to preload documents from tertiary storage was made by Kraiss [13] This work is probably most related to our work because it uses a ....

K.S. Grimsrud, J.K. Archibald, and B.E. Nelson. Multiple Prefetch Adaptive Disk Caching. IEEE Knowledge and Data Engineering, 5(1):88--103, February 1993.

....supports command queuing that permits a disk controller to implement an intelligent optimization, such as seek ordering. Examples of host based prefetching include file system caches that prefetch a large portion of the file being read, and query optimizers in database applications. Thus, [Grimsrud93 ] proposed a caching strategy for a host based segmented cache. The strategy includes maintaining a record of the order of past disk accesses for predicting future access sequences. Our study has another focus. We concentrate mostly on a disk based segmented cache and on an analyses of disk load ....

K. Grimsrud, J. Archibald, and B. Nelson. Multiple Prefetch Adaptive Disk Caching. IEEE Trans. on Knowledge and Data Engineering, 5(1) (1993), pp. 88-103.

....buffers based on the application access context, and proposes customized replacement algorithms for each state. Prefetching is performed in the MMFS using the application hints regarding the rate and direction of retrieval. Intelligent prefetching schemes for general applications are presented in [18, 59, 65]. Various replacement strategies for file servers are analyzed and simulated in [36, 73] Their simulation bears out their conjecture that file servers and clients should use different replacement policies. Adapting their logic into the MMFS model, it is perhaps a good idea not to implement the ....

Grimsrud, K. et al. Multiple prefetch adaptive disk caching. IEEE Transactions on Knowledge and Data Engineering, 5(1):88--103, February 1993.

No context found.

K. Grimsrud, J. Archibald, and B. Nelson, Multiple prefetch adaptive disk caching, IEEE Transactions on Knowledge and Data Engineering, February 1993.

No context found.

K.S. Grimsrud, J.K. Archibald, and B.E. Nelson, "Multiple Prefetch Adaptive Disk Caching" IEEE Transactions on Knowledge and Data Engineering, Feb. 1993, pp. 88103. 29

No context found.

K. Grimsrud, J. Archibald, and B. Nelson, "Multiple prefetch adaptive disk caching", IEEE Transactions on Knowledge and Data Engineering, February 1993.

No context found.

K. Grimsrud, J. Archibald, and B. Nelson, Multiple prefetch adaptive disk caching, IEEE Transactions on Knowledge and Data Engineering, February 1993.

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