Failure correction techniques for large disk arrays

Garth A. Gibson; Lisa Hellerstein; Richard M. Karp; Randy H. Katz; David A. Patterson

doi:10.1109/9780470544839.ch11

Failure correction techniques for large disk arrays

Garth A. Gibson, Lisa Hellerstein, Richard M. Karp, Randy H. Katz, David A. Patterson

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Abstract

The ever increasing need for I/O bandwidth will be met with ever larger arrays of disks. These arrays require redundancy to protect against data loss. This paper examines alternative choices for encodings, or codes, that reliably store information in disk arrays. Codes are selected to maximize mean time to data loss or minimize disks containing redundant data, but are all constrained to minimize performance penalties associated with updating information or recovering from catastrophic disk failures. We show codes that give highly reliable data storage with low redundant data overhead for arrays of 1000 information disks.

Original language	English (US)
Title of host publication	High Performance Mass Storage and Parallel I/O
Subtitle of host publication	Technologies and Applications
Publisher	Wiley-IEEE Press
Pages	149-160
Number of pages	12
ISBN (Electronic)	9780470544839
ISBN (Print)	0471208094, 9780471208099
DOIs	https://doi.org/10.1109/9780470544839.ch11
State	Published - Jan 1 2001

Keywords

Arrays
Encoding
Maintenance engineering
Measurement
Parity check codes
Redundancy

ASJC Scopus subject areas

Computer Science(all)

Access to Document

10.1109/9780470544839.ch11

Fingerprint Dive into the research topics of 'Failure correction techniques for large disk arrays'. Together they form a unique fingerprint.

Cite this

@inbook{7c5857a967fd48658b8ce89572c0d6be,

title = "Failure correction techniques for large disk arrays",

abstract = "The ever increasing need for I/O bandwidth will be met with ever larger arrays of disks. These arrays require redundancy to protect against data loss. This paper examines alternative choices for encodings, or codes, that reliably store information in disk arrays. Codes are selected to maximize mean time to data loss or minimize disks containing redundant data, but are all constrained to minimize performance penalties associated with updating information or recovering from catastrophic disk failures. We show codes that give highly reliable data storage with low redundant data overhead for arrays of 1000 information disks.",

keywords = "Arrays, Encoding, Maintenance engineering, Measurement, Parity check codes, Redundancy",

author = "Gibson, {Garth A.} and Lisa Hellerstein and Karp, {Richard M.} and Katz, {Randy H.} and Patterson, {David A.}",

year = "2001",

month = jan,

day = "1",

doi = "10.1109/9780470544839.ch11",

language = "English (US)",

isbn = "0471208094",

pages = "149--160",

booktitle = "High Performance Mass Storage and Parallel I/O",

publisher = "Wiley-IEEE Press",

}

TY - CHAP

T1 - Failure correction techniques for large disk arrays

AU - Gibson, Garth A.

AU - Hellerstein, Lisa

AU - Karp, Richard M.

AU - Katz, Randy H.

AU - Patterson, David A.

PY - 2001/1/1

Y1 - 2001/1/1

N2 - The ever increasing need for I/O bandwidth will be met with ever larger arrays of disks. These arrays require redundancy to protect against data loss. This paper examines alternative choices for encodings, or codes, that reliably store information in disk arrays. Codes are selected to maximize mean time to data loss or minimize disks containing redundant data, but are all constrained to minimize performance penalties associated with updating information or recovering from catastrophic disk failures. We show codes that give highly reliable data storage with low redundant data overhead for arrays of 1000 information disks.

AB - The ever increasing need for I/O bandwidth will be met with ever larger arrays of disks. These arrays require redundancy to protect against data loss. This paper examines alternative choices for encodings, or codes, that reliably store information in disk arrays. Codes are selected to maximize mean time to data loss or minimize disks containing redundant data, but are all constrained to minimize performance penalties associated with updating information or recovering from catastrophic disk failures. We show codes that give highly reliable data storage with low redundant data overhead for arrays of 1000 information disks.

KW - Arrays

KW - Encoding

KW - Maintenance engineering

KW - Measurement

KW - Parity check codes

KW - Redundancy

UR - http://www.scopus.com/inward/record.url?scp=85036531056&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85036531056&partnerID=8YFLogxK

U2 - 10.1109/9780470544839.ch11

DO - 10.1109/9780470544839.ch11

M3 - Chapter

AN - SCOPUS:85036531056

SN - 0471208094

SN - 9780471208099

SP - 149

EP - 160

BT - High Performance Mass Storage and Parallel I/O

PB - Wiley-IEEE Press

ER -