Mixed Ordering RAID Stripe Resilience for SSD Zones

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current data storage systems face challenges in maintaining reliability due to variations in error rates across different storage media, which can overwhelm error correction mechanisms, especially when dealing with high error rates during data access.

Innovation Solution

Implementing a mixed ordering scheme for RAID stripes across heterogeneous solid-state memory, where some pages are used for dummy data or extra parity, to enhance read reliability and manage variations in storage media reliability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If error correction code (ECC) data and RAID technology are used to safeguard against failures, then reliability is improved, but the system can be overwhelmed when the number of errors exceeds error correction capability

Engineering Contradiction:
Improvestorage system reliabilityVSAvoiderror correction system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the storage system into multiple independent RAID groups, each with its own error correction capabilities. By dividing the storage array into smaller manageable units, the system can isolate errors to specific segments rather than having errors propagate system-wide, allowing error correction to remain effective even when many errors occur across the entire system.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements different error correction strategies for different RAID groups based on their specific characteristics and error patterns. Each RAID group can be configured with appropriate ECC levels and redundancy schemes tailored to its workload and failure modes, optimizing reliability without uniformly increasing complexity across the entire system.

Inventive Principle:
Principle #3Local quality

2Quantity of substance

If denser storage media and larger storage capacity are implemented, then storage capacity is improved, but failure mechanisms and error rates increase

Engineering Contradiction:
Improvestorage capacityVSAvoidstorage media reliability
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent divides large-capacity storage systems into multiple smaller RAID groups, each managing a portion of the total storage capacity. This segmentation allows error correction to be applied effectively at the group level, preventing errors in dense storage media from overwhelming the entire system while still providing access to large total capacity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent dynamically adjusts error correction parameters and redundancy levels based on the specific characteristics of the storage media being used. As storage density increases and error rates change, the system can modify ECC strength, RAID configuration, and protection levels to maintain reliability while maximizing usable capacity.

Inventive Principle:
Principle #35Parameter changes

3Device complexity

If uniform ordering of RAID stripes is used, then simplicity is maintained, but variations in storage media reliability cause read errors

Engineering Contradiction:
ImproveRAID stripe ordering complexityVSAvoidread operation reliability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent implements mixed ordering of RAID stripes where different portions of the storage array use different ordering strategies tailored to the specific reliability characteristics of each storage media type. This allows the system to optimize read reliability for each media type (e.g., SSD vs. HDD) without imposing complex ordering requirements across the entire uniform system.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent deliberately introduces asymmetric ordering patterns in RAID stripes that account for the different failure modes and access patterns of heterogeneous storage media. Rather than using symmetric uniform ordering, the system employs asymmetric arrangements that place data and parity strips in positions optimized for each media type's reliability characteristics, improving read operations without requiring complex coordination.

Inventive Principle:
Principle #4Asymmetry

Data Source

PatentUS12056365B2Resiliency for a storage system
Publication Date: 2024.08.06 PURE STORAGE INC
  • US12056365B2 patent drawing
  • US12056365B2 patent drawing
  • US12056365B2 patent drawing

AI summary

A storage system has a resiliency scheme to enhance storage system performance. The storage system composes a RAID stripe. The storage system mixes an ordering of portions of the RAID stripe, based on reliability differences across zones of the solid-state memory. Each zone of the solid state memory corresponds to a type of solid state memory. The storage system writes the mixed ordering RAID stripe across the solid-state memory.