Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

A simple low cost IDE RAID 1 Autonomic implementation

IP.com Disclosure Number: IPCOM000019259D
Original Publication Date: 2003-Sep-08
Included in the Prior Art Database: 2003-Sep-08
Document File: 2 page(s) / 56K

Publishing Venue

IBM

Abstract

Many of todays popular OS's provide native support for software mirroring (Raid 1). Although software mirroring support can impact disk write performance since two write operations are needed at the system level to insure that the data gets written to both disks, some customers or market segments find this performance degradation acceptable given the added insurance they get with the redundancy. The problem with software mirroring is that an overt action is needed to continue to make a system function (or boot) in the event that the primary HDD on the ATA interface has had a electronics failure. This overt action may involve the IT Admin swapping jumpers on the ATA Slave device to make it a Master. This is required due to the design of the ATA interface which requires the presence of a master device in order for a slave device to operate properly. What is needed is an autonomic solution which detects a defective ATA Master device, disables it from being the master and automatically reconfigures the slave device to become the master device. In this manner, the system practicing the present invention, recovers to the backup copy of the HDD, thus realizing minimal down time. POST needs to know the state of the cable, try primary disk, get time out, set flags in system and generate alert (could be via the RSA), flip cable lines, issue IDE reset on IDE port, wait for sync time as per the ATA spec, retry boot on next disk, boot to OS, flag offline disk at OS agent level.

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 78% of the total text.

Page 1 of 2

A simple low cost IDE RAID 1 Autonomic implementation

       A system with IDE hardfiles employing Software IDE RAID does not allow for autonomic system recovery if the Master ATA disk fails. Manual intervention is required by a user to reactivate or boot to the backup drive. The present invention allows for unassisted recovery to the back disk when detecting a disk error either in the MBR or in the drive electronics. When an error or failure is detected a system can automatically reassign the HDD's address by programmatically controlling which drive is the ATA Master and Slave using the cable of the present invention. This is accomplished by programming the signals on the cable select lines.

     The ATA architecture allows for two devices to be connected to the existing IDE cable, Device 0 and Device 1. The Device 0 (e.g. master device) signal on the connecting ATA cable is terminated at the Device 0 connector. The signal does is not propagated to the connector for Device 1 on the existing cable. In order for Device 1(e.g. slave device) to become a master, external insertion of a master control signal will be used when the master device fails. This control line will be inserted at the Device 1 side of the "Open" connection and it will be controlled by the system practicing the present invention, when it detects failure of the master device.

Ref. document page 21 in the ATA Spec

1

[This page contains 2 pictures or other non-text objects]

Page 2 of 2


1.


2.


3.


4.


5.

Host boot...