RAID Cache Stripe Area for Data Consistency Recovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional RAID systems face reliability issues when data consistency is broken due to failures in the control module, leading to potential data garbling during recovery processes, especially when parity cannot be saved or recalculated effectively.

Innovation Solution

A storage system with a cache control unit, RAID control unit, write-back processing unit, and recovery processing unit that manages data in page and strip area units, reserves a cache stripe area for correct data, and stages new data using exclusive OR operations to prevent data garbling and ensure reliable recovery.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional write-back process is used in RAID 5 system, then data can be written to storage devices, but data consistency may be broken when control module fails or multiple disk errors occur

Engineering Contradiction:

Improvedata consistencyVSAvoiddata garbling

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent applies preliminary action by reserving a cache stripe area in advance before write-back operations. This reserved area stores correct data that can be used for recovery if consistency is broken. The system prepares the recovery mechanism beforehand rather than attempting to fix consistency after failure occurs.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent uses the cache stripe area as an intermediary between the failed disk devices and the recovery process. This intermediary stores correct data that mediates the recovery operation, allowing the system to restore consistency without directly accessing potentially corrupted data on failed disks.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If parity buffer area is released after write-back, then cache memory can be reused, but recovery cannot be performed when consistency is broken

Engineering Contradiction:

Improvecache memory utilizationVSAvoiderror recovery capability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent extracts the recovery capability from the parity buffer area by storing correct data in the reserved cache stripe area. This separation allows the parity buffer to be released for cache reuse while the cache stripe area maintains the recovery function independently.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs preliminary action by reserving the cache stripe area and storing correct data there before write-back operations complete. This ensures recovery data is available in advance, allowing the parity buffer to be released without compromising recovery capability.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If control module fails and backup control module takes over, then system availability is maintained, but parity cannot be saved or recalculated effectively

Engineering Contradiction:

Improvesystem availabilityVSAvoidparity information

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The patent uses copying by storing correct data in the reserved cache stripe area as a copy that can be used for recovery. This copy of correct data replaces the need to recalculate parity from potentially corrupted source data, preserving parity information even when control modules fail.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The system performs preliminary action by reserving the cache stripe area and populating it with correct data before control module failure occurs. This ensures parity information is preserved in advance, allowing the backup control module to perform recovery without needing to recalculate parity.

Inventive Principle:

Principle #10Preliminary action

4Reliability

If Regeneration Read is executed from disk devices with broken consistency, then data can be recovered through XOR operation, but garbled data may be restored

Engineering Contradiction:

Improvedata recoveryVSAvoiddata accuracy

Core Design Contradiction:

ReliabilityVSManufacturing precision

Solution Approach 1:

The patent uses the cache stripe area as an intermediary that provides accurate source data for recovery operations. By reading correct data from this intermediary rather than directly from corrupted disk devices, the XOR operation produces accurate recovered data without garbling.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent converts the potential harm of corrupted disk data into a benefit by using the reserved cache stripe area with correct data as the source for recovery. The system transforms the problem of corrupted source data into an advantage by having pre-stored correct data available for the XOR recovery operation.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Data Source

PatentUS7370148B2Storage system, control method thereof, and program

Publication Date: 2008.05.06 FUJITSU LTD

AI summary

When a write-back request for writing back new data in a cache memory to disk devices forming a redundant configuration of RAID is generated, a write-back processing unit, reserves a parity buffer area in the cache memory, generates new parity, and then, writes the new data and the new parity to the corresponding disk devices. When an error in which the consistency of RAID is broken occurs in a plurality of the disk devices upon write performed by the write-back processing unit, a recovery processing unit reserves, in the cache memory, a cache stripe area storing data of the entire stripe including the new data which is to be written back, and causes the cache control unit to manage it.