Browse Prior Art Database

Temporary RAID System Sparing to Low Cost Devices

IP.com Disclosure Number: IPCOM000032772D
Original Publication Date: 2004-Nov-11
Included in the Prior Art Database: 2004-Nov-11
Document File: 4 page(s) / 138K

Publishing Venue

IBM

Abstract

Direct access storage device (DASD) storage systems that manage multiple RAID arrays typically limit their underlying storage devices to a particular protocol class/type. DASD RAID controllers that use global sparing need methods/rules to determine what storage devices are spare candidates for particular RAID arrays. Allowing storage device intermix that goes beyond different size and speed (i.e.; FC-AL, SATA, SAS) within the same global sparing domain can create reliability and performance issues with RAID array sparing scenarios. Maintaining all enterprise class devices for a particular system requires that a customer bear the cost of enterprise storage devices for all devices including spare devices that are not being actively used.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 4

Temporary RAID System Sparing to Low Cost Devices

Disclosed is a system and method to reduce the total cost of ownership where it is possible to exploit the capability to intermix storage device classes (server class versus desktop class) for use global spares. To provide the most economical configurations a user can implement a system configuration such that they create all of the RAID Array members (All data drives and parity drives) with enterprise class storage devices (i.e.; FC-AL) and implement all global hot spares will a lower cost storage device (i.e. SATA). This configuration allows for a sparing action to use the lower cost storage device temporarily until the device that has failed (cause of sparing action) is replaced with an equivalent class of storage device (FC-AL). When the repair activity is complete and the new repaired (FC-AL) storage device is available a copy back function would be utilized to restore the RAID Array members to the original configuration (All data and parity drives enterprise class) and the storage device that was used for the sparing action will return to its state as a global hot spare. An analogy to this activity is the "space saver" spare tires that are designed for temporary repairs of flat tires for automobiles.

Normal Running Configuration

RAID 5 Array (3+P)

RAID 10 Array (2x2)

RAID 5 Array (7+P)

SATA DDM

FC-AL DDM

For a sparing activity the configuration changes as below.

1

Page 2 of 4

Sparing Action Configuration

RAID 5 Array (7+P)

X

RAID 5 Array (3+P)

RAID 10 Array (2x2)

SATA DDM

FC-AL DDM

Once the failing device has been replaced and is eligible to an Array Member the copy back function is used.

2

[This page contains 1 picture or other non-text object]

Page 3 of 4

Copy Back Repair Action Configuration

RAID 5 Array (3+P)

RAID 10 Array (2x2)

RAID 5 Array (7+P)

SATA DDM

FC-AL DDM

The configuration returns to the normal configuration above.

The performance expectations of the array in this temporary state need to account for the performance characteristics of the new array members tha...