de.mpg.escidoc.pubman.appbase.FacesBean
English
 
Help Guide Disclaimer Contact us Login
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Report

Fast concurrent access to parallel disks

MPS-Authors
http://pubman.mpdl.mpg.de/cone/persons/resource/persons45344

Sanders,  Peter
Algorithms and Complexity, MPI for Informatics, Max Planck Society;

http://pubman.mpdl.mpg.de/cone/persons/resource/persons44368

Egner,  Sebastian
Algorithms and Complexity, MPI for Informatics, Max Planck Society;

Locator
There are no locators available
Fulltext (public)

1999-1-003
(Any fulltext), 11KB

Supplementary Material (public)
There is no public supplementary material available
Citation

Sanders, P., Egner, S., & Korst, J.(1999). Fast concurrent access to parallel disks (MPI-I-1999-1-003). Saarbrücken: Max-Planck-Institut für Informatik.


Cite as: http://hdl.handle.net/11858/00-001M-0000-0014-6F94-0
Abstract
High performance applications involving large data sets require the efficient and flexible use of multiple disks. In an external memory machine with D parallel, independent disks, only one block can be accessed on each disk in one I/O step. This restriction leads to a load balancing problem that is perhaps the main inhibitor for the efficient adaptation of single-disk external memory algorithms to multiple disks. We show how this problem can be solved efficiently by using randomization and redundancy. A buffer of O(D) blocks suffices to support efficient writing of arbitrary blocks if blocks are distributed uniformly at random to the disks (e.g., by hashing). If two randomly allocated copies of each block exist, N arbitrary blocks can be read within ceiling(N/D)+1 I/O steps with high probability. The redundancy can be further reduced from 2 to 1+1/r for any integer r. From the point of view of external memory models, these results rehabilitate Aggarwal and Vitter's "single-disk multi-head" model that allows access to D arbitrary blocks in each I/O step. This powerful model can be emulated on the physically more realistic independent disk model with small constant overhead factors. Parallel disk external memory algorithms can therefore be developed in the multi-head model first. The emulation result can then be applied directly or further refinements can be added.