Non-Volatile Storage Verification via Segmented Read-Only Testing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Data storage systems face challenges in verifying the integrity of unneeded data in non-volatile storage after a power cycle, risking data loss or corruption without proper testing mechanisms.

Innovation Solution

Implementing a read-only test to verify unneeded data in non-volatile storage by detecting data parity and ECC uncorrectables, ensuring data integrity without affecting needed data, and conducting a write and readback test if the data is correctly preserved.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the non volatile storage is put into self refresh mode to preserve data during power loss, then data is maintained for a period of time, but the storage cannot be verified for integrity without risking data loss

Engineering Contradiction:

Improvedata preservationVSAvoidverification process

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The verification process is segmented into two distinct paths: one for preserved data (read-only tests) and one for non-preserved data (write and readback tests). This segmentation allows verification without risking data loss in preserved regions while maintaining thorough testing for non-preserved regions.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

For preserved data, only partial verification (read-only tests) is performed instead of complete verification (write and readback tests). This partial action is sufficient to verify integrity without risking data loss, whereas excessive action would attempt to write test patterns that could corrupt preserved data.

Inventive Principle:

Principle #16Partial or excessive action

2Measurement precision

If the IML sequence tests the non volatile storage by writing and reading back test patterns, then defective memory is detected, but data that is needed may be lost

Engineering Contradiction:

Improvememory defect detectionVSAvoiddata integrity

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

Different verification methods are applied to different regions of non volatile storage based on their preservation status. Preserved data regions receive read-only verification to maintain integrity, while non-preserved regions receive full write and readback testing to detect defects. This local differentiation resolves the contradiction between thorough testing and data protection.

Inventive Principle:

Principle #3Local quality

3Productivity

If the non volatile storage is not preserved or contains unneeded data, then the IML sequence can conduct full testing, but preserved needed data may not be verified

Engineering Contradiction:

Improvetesting throughputVSAvoiddata verification

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system uses feedback from the preservation status indicator to determine the appropriate verification method. This feedback mechanism ensures that preserved data is verified through read-only tests while non-preserved data undergoes full testing, optimizing both verification reliability and processing efficiency.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS7958432B2Verification of non volatile storage storing preserved unneeded data

Publication Date: 2011.06.07 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Non volatile storage may be employed to temporarily store data which is destaged to data storage drives. The non volatile storage is configured to preserve the data through a power outage. Some data may be preserved, but is not needed, such as the result of a failover to another non volatile storage. This unneeded data is tested to verify the non volatile storage by indicating whether the data survived the power cycle from full power to self refresh mode battery power to full power, without risking the loss of data that is needed.