RAID Reconstruction Using Reparity Bitmap Tracking

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

RAID systems face performance overhead and reliability issues during data reconstruction due to concurrent operation with user I/O requests, leading to prolonged reconstruction times and increased vulnerability to secondary disk failures.

Innovation Solution

Implementing a method to selectively limit the data set considered in RAID operations by tracking the highest disk block number (DBN) written and using a reparity bitmap to reduce the number of stripes that need to be reconstructed, allowing for partial rather than full disk reconstruction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If full disk reconstruction is performed, then data reliability is improved, but reconstruction time increases and system vulnerability to secondary failures increases

Engineering Contradiction:

Improvedata reliabilityVSAvoidreconstruction time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent applies partial action by performing reconstruction only on the portion of the disk that contains valid data. The system tracks the highest disk block number (DBN) written by the file system and reconstructs only up to that point, rather than performing full disk reconstruction. This reduces reconstruction time and system vulnerability while maintaining data reliability for all actually used data blocks.

Inventive Principle:

Principle #16Partial or excessive action

2Adaptability or versatility

If concurrent user I/O requests are handled during reconstruction, then service continuity is maintained, but reconstruction performance degrades

Engineering Contradiction:

Improveservice continuityVSAvoidreconstruction performance

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent extracts the harmful effect of concurrent I/O during reconstruction by using a surrogate RAID to handle user write requests. The surrogate RAID accepts writes to the failed disk's data blocks, allowing the reconstruction process to proceed without interference from user I/O requests. The extracted write data is later reclaimed to the original RAID after reconstruction completes.

Inventive Principle:

Principle #2Taking out (Extraction)

3Reliability

If full disk capacity is considered in RAID operations, then complete data protection is achieved, but operation time increases

Engineering Contradiction:

Improvedata protection completenessVSAvoidoperation time

Core Design Contradiction:

ReliabilityVSDuration of action of moving object

Solution Approach 1:

The patent applies preliminary action by pre-tracking the highest DBN written by the file system using a reparity bitmap before reconstruction is needed. This bitmap records which disk blocks contain valid data, allowing the system to determine the exact scope of reconstruction needed in advance. This eliminates the need to scan or consider the entire disk capacity during reconstruction operations.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11537330B2Selectively improving raid operations latency

Publication Date: 2022.12.27 NETAPP INC

AI summary

Presented herein are methods, non-transitory computer readable media, and devices for selectively limiting the amount of data in a file system, which include: determining a reparity bit value for a write disk block range, wherein the reparity bit is configured to track a number of writes in progress to a stripe range; determining the reparity bit value; updating a threshold written disk block number as a highest disk block number of the reparity bit value; and initiating a RAID operation until it reaches the threshold written disk block number, wherein the threshold written disk block number comprises a maximum written disk block number representing the last disk block number written.