Memory Parity Update Verification for RAID Data Integrity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing memory systems face challenges in ensuring fault tolerance and data integrity during updates, particularly in RAID configurations, where parity generation and verification processes are inefficient and prone to errors.

Innovation Solution

A memory system and method that generates and verifies parities using XOR calculations and Galois field operations to ensure accurate parity updates, allowing for reliable data reconstruction and verification in RAID-5 and RAID-6 configurations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional parity generation methods are used in RAID configurations, then data redundancy is achieved, but data integrity during updates cannot be ensured due to inefficiency and error-proneness

Engineering Contradiction:
Improvedata integrityVSAvoidparity generation efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by generating intermediate parity data before the actual update operation. The controller calculates intermediate parity based on old data and new data, then uses this intermediate parity to verify the correctness of the update process. This preliminary calculation enables verification before completion, ensuring data integrity while maintaining efficiency through optimized XOR operations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback mechanisms by continuously verifying parity data during the update process. The controller reads previously stored parity, generates new parity based on updated data, and compares the generated parity with expected values. This feedback loop detects errors early and prevents corrupted data from being committed, thereby ensuring data integrity without significantly impacting overall update efficiency.

Inventive Principle:
Principle #23Feedback

2Reliability

If parity verification is performed during updates, then data correctness is improved, but processing time increases

Engineering Contradiction:
Improvefault toleranceVSAvoidupdate processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies partial action by performing verification on only the changed portions of data rather than the entire dataset. When updating specific data blocks, the controller generates and verifies parity only for the affected stripes, leaving unchanged data unverified. This selective verification approach maintains fault tolerance for updated data while minimizing the time overhead associated with comprehensive verification.

Inventive Principle:
Principle #16Partial or excessive action

3Reliability

If multiple parities are stored for RAID-6, then fault tolerance is improved, but system complexity increases

Engineering Contradiction:
Improvefault toleranceVSAvoidparity management complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the parity management process into distinct stages: generating first parity based on original data, generating second parity based on updated data and first parity, and verifying each parity separately. This segmented approach to handling multiple parities in RAID-6 reduces management complexity by treating each parity generation and verification as an independent, manageable task rather than a monolithic complex process.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4703886A1Memory system and method of operating memory systems
Publication Date: 2026.03.04 KIOXIA CORP
  • EP4703886A1 patent drawingFigure 1
  • EP4703886A1 patent drawingFigure 2
  • EP4703886A1 patent drawingFigure 3

AI summary

According to one embodiment, a memory system includes an interface circuit (11) communicable with a host (2) and a plurality of external memory systems (3-1, 3-2, 3-3, 3-4, 3-5), a nonvolatile memory (5), and a controller (14). The plurality of external memory systems (3-1, 3-2, 3-3, 3-4, 3-5) include at least a first memory system (3-3) and a second memory system (3-2). The controller (11) is configured to generate first parity (Old Q) based on first data (Old D2) stored in the first memory system (3-3) and second data (Old D1) stored in the second memory system (3-2), write the first parity (Old Q) to the nonvolatile memory (5), receive third data (New D2) that is updated data of the first data (Old D2) from the host (2) via the interface circuit (11), receive the first data (Old D2) from the first memory system (3-3) via the interface circuit (11), read the first parity (Old Q) from the nonvolatile memory (5), generate second parity (New Q) based on the first parity (Old Q), the first data (Old D2), and the third data (New D2), generate first rebuilt data (Rebuild D2(Q)) based on the second parity (New Q), the first parity (Old Q), and the first data (Old D2), and compare the first rebuilt data (Rebuild D2(Q)) with the third data (New D2).