Loading...
Thumbnail Image
Publication

Duality Between Prefetching and Queued Writing with Parallel Disks

Hutchinson, David A.
Sanders, Peter
Vitter, Jeffrey Scott
Citations
Altmetric:
Abstract
Parallel disks promise to be a cost effective means for achieving high bandwidth in applications involving massive data sets, but algorithms for parallel disks can be difficult to devise. To combat this problem, we define a useful and natural duality between writing to parallel disks and the seemingly more difficult problem of prefetching. We first explore this duality for applications involving read-once accesses using parallel disks. We get a simple linear time algorithm for computing optimal prefetch schedules and analyze the efficiency of the resulting schedules for randomly placed data and for arbitrary interleaved accesses to striped sequences. Duality also provides an optimal schedule for prefetching plus caching, where blocks can be accessed multiple times. Another application of this duality gives us the first parallel disk sorting algorithms that are provably optimal up to lower-order terms. One of these algorithms is a simple and practical variant of multiway mergesort, addressing a question that had been open for some time.
Description
This is the published version, made available with the permission of the publisher. Copyright © 2005 Society for Industrial and Applied Mathematics.
Date
2006-07-27
Journal Title
Journal ISSN
Volume Title
Publisher
Society for Industrial and Applied Mathematics
Research Projects
Organizational Units
Journal Issue
Keywords
Caching, External memory sorting, Load balancing, Lower bound, Prefetching, Randomized algorithm
Citation
Hutchinson, David A., Peter Sanders, and Jeffrey Scott Vitter. "Duality Between Prefetching and Queued Writing with Parallel Disks." SIAM J. Comput. SIAM Journal on Computing 34.6 (2005): 1443-463. DOI:10.1137/S0097539703431573
Embedded videos