Asymmetrical RAID 6 Parity Layout for Low-Latency Disk Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional RAID 6 schemes are complex and incur high read latency during disk failures, requiring extensive reading of data stripes to recover a single data block, which complicates the recovery process.
Innovation Solution
The implementation of a simplified RAID 6 scheme that reduces read latency by using a unique configuration with asymmetrical data and parity columns, allowing for recovery without reading the entire data stripe and utilizing diagonal and row parity blocks efficiently to minimize read operations, and includes a disk addition unit to manage new disk additions without disrupting parity distribution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional RAID 6 schemes are used, then data redundancy and reliability are maintained, but read latency increases significantly during disk failures due to extensive reading of data stripes required for recovery
Solution Approach 1:
The patent segments the parity data into two distinct types: column parity data stored in dedicated column parity blocks and diagonal parity data stored in dedicated diagonal parity blocks. This segmentation allows the system to perform targeted recovery operations by reading only the specific parity blocks needed for the failed disk, rather than reading entire data stripes, thereby reducing read latency while maintaining data redundancy
Solution Approach 2:
The patent introduces a diagonal parity dimension in addition to the traditional column parity structure. By organizing parity data in two dimensions (column and diagonal), the system can recover from single or double disk failures by reading only the necessary parity blocks along the failed disk's column and diagonal, significantly reducing the read operations required compared to traditional stripe-based recovery
2Reliability
If traditional RAID 6 schemes are used, then data protection is achieved, but the complexity of the recovery process increases due to extensive reading of data stripes
Solution Approach 1:
The patent segments the recovery process into distinct phases based on the type of failure (single disk or double disk) and the location of failed disks. The segmented parity block structure enables the system to follow simplified recovery paths: for single disk failures, read column parity blocks; for double disk failures, read both column and diagonal parity blocks. This segmentation reduces recovery process complexity compared to the monolithic stripe-based recovery approach
3Loss of time
If asymmetrical data and parity column configuration is used, then read latency is reduced during recovery, but the device structure becomes more complex
Solution Approach 1:
The patent employs an asymmetrical configuration where data blocks and parity blocks are organized in non-uniform patterns with different numbers of data rows and columns. This asymmetry optimizes the storage structure for efficient recovery operations by aligning the data layout with the parity block structure, enabling faster recovery while the managing unit handles the complexity of the asymmetrical arrangement
4Adaptability or versatility
If new disks are added to the RAID system, then storage capacity is increased, but parity distribution must be maintained without disrupting existing data
Solution Approach 1:
The patent implements a universal managing unit that handles multiple functions: initial parity distribution, dynamic disk addition, and recovery operations. This universal controller can adapt to different RAID configurations and dynamically manage parity distribution when disks are added, maintaining system integrity while providing flexibility for expansion without requiring complex manual reconfiguration
Data Source
AI summary
A system, computer program product, and computer-executable method of implementing a redundant array of independent disk (RAID) system wherein the RAID, the computer-executable method comprises storing data storage blocks arranged in a first plurality of data rows and a second plurality of data columns and storing parity data in defined parity blocks, wherein a portion of the defined parity blocks include column parity data.