S. Akyurek and K. Salem. Adaptive block rearrangement. CS-- TR--2854. University of Maryland, February 1992.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....storage, replace current devices with faster ones, or upgrade the host adapter to one that supports caching and command queuing. Change the location of les on the storage devices. For example, move frequently accessed disk les to the middle tracks to reduce the average seek distance [62]. Some storage systems dynamically monitor their own performance, and modify their behavior based on the measurements. For example, software in the EMC disk array controller maintains hundreds of counters that are used by the controller to tune its execution strategies under speci c workloads. ....

S. Akyurek and K. Salem, \Adaptive block rearrangement," in Proceedings of the Ninth International Conference on Data Engineering, (Vienna, Austria), pp. 182{ 189, IEEE Computer Society Press, Apr. 1993.

....of small files from the same directory in adjacent blocks that are accessed as a unit [13] A second approach is to rearrange data that is already on disk. Researchers have explored reducing seek times by rearranging on disk data such that frequently accessed data is near the middle of the disk [66, 55, 44, 1]. This research exploits the observation that, in general, there is a small subset of the data on a disk which contains the data that will be specified by most of the requests to that disk. applications that must fetch data from disk to spend an increasing proportion of their execution time ....

....output (as, for example, in the program shown in Figure 3. 1) Another likely scenario is struct fileheader int offsetofAdata; int sizeofAdata; int offsetofKdata; int sizeofKdata; void main(int argc, char argv) int fd; struct fileheader myfh; char buf; fd = open(argv[1], r ) read(fd, myfh, sizeof(struct fileheader) lseek(fd, myfd.offsetofKdata, SEEK SET) buf = malloc(myfd.sizeofKdata) read(fd, buf, myfd.sizeofKdata) process Kdata(buf) Figure 3.5: A simple program used to illustrate how data dependencies can make it unlikely that any automatic ....

S. Akyurek and K. Salem. Adaptive block rearrangement. In Proceedings of the IEEE International Conference on Data Engineering, April 1993.

....prevents a block from migrating to different disks over time. The desirable data striping property of the original RAID design, therefore, can be preserved to ensure efficient data fetching. During the relocation, frequently referenced blocks can be moved to the hot region to reduce seek time [1, 7]. Another advantage of Multibuffering is due to the fact that the buffer space also serves as a cache to reduce the I O activities. We will discuss the buffer management policy in details later. 2 a non volatile storage which is used in delayed writing schemes including DPSR to rebuild data in ....

S. AKyuret and K. Salem. Adaptive block rearrangement. In Proc. of the Nineth ICDE, pages 182--189, April 1993.

.... HL1 Figure 1: The Wavelet Decomposition Process Hence the Euclidean distance between a pair of images gives a quantitative measure of the degree of similarity of their image contents [AMEM95] 3 Related Work Data placement has been an important research topic in the context of database systems [AS95, BG88, CABK88] The technique of striping data across multiple disks to overcome the I O bottleneck has been extensively studied [PGK88, CP90] The improvements in performance are essentially because large requests are serviced in parallel by multiple disks and multiple small requests are handled ....

....declustering method, called Coordinate Modulo Distribution (CMD) is developed in [LSR92] When the structure of the data or the access patterns are not well understood, other approaches have been taken to improve I O performance. In these cases data is generally viewed as independent files. In [AS95] a technique which dynamically copies blocks that are frequently accessed to a reserved space in the center of the disk to reduce seek delays is described. The notion of Disk Cooling is introduced in [SWZ94] where data is dynamically redistributed based on the access patterns to yield higher ....

S. Akyurek and K. Salem. Adaptive block rearrangement. ACM Trans. on Comp. Systems, 13(2):89--121, May 1995.

....secondary storage devices has come to be known as the I O bottleneck. One of the major components of disk access times is seek time, the time required to move the drive s read write heads to the required data. By reducing seek time, the performance of a disk can be improved. In a recent paper [Akyurek 93] we introduced an adaptive block rearrangement technique which reduces disk seek times. Under this technique, a small number of frequently referenced data blocks are copied from their original locations to a reserved space near the middle of the disk. The term adaptive means that no advance ....

....that does not appear on the list is referenced, a replacement heuristic is used to make room for it on the list. In our experiments, the analyzer maintained a list of several thousand reference counts, enough so that replacement was rarely necessary. However, experiments reported in [Salem 92, Akyurek 93, Salem 93] indicate that if space is limited, this technique can still generate very accurate guesses using much shorter lists. Given a list of block reference frequencies, the block arranger selects the hottest blocks from the list and determines where to place them in the reserved region. For ....

[Article contains additional citation context not shown here]

Akyurek, Sedat, Kenneth Salem, "Adaptive Block Rearrangement," Proceedings of Ninth International Conference on Data Engineering, Vienna, Austria, April 1993.

No context found.

S. Akyurek and K. Salem. Adaptive block rearrangement. CS-- TR--2854. University of Maryland, February 1992.

No context found.

S. Akyurek and K. Salem, `Adaptive Block Rearrangement', ACM Transactions on Computer Systems, 13, (2), 89-121 (1995).

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