Vault Failure Abatement for Dispersed Storage Recovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Dispersed storage networks face challenges in managing and recovering data when storage units fail, especially when these failures affect multiple vaults, as existing systems lack efficient methods to abate failures and maintain data integrity across geographically distributed storage units.

Innovation Solution

The implementation of a dispersed storage network with a managing unit and integrity processing unit that performs error encoding and decoding using Cauchy Reed-Solomon encoding, allowing for data redundancy and slice management, enabling the network to tolerate storage unit failures by rebuilding 'bad' or missing encoded data slices and migrating data from failed units to non-failed ones.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data is dispersed across multiple storage units to improve reliability, then system resilience improves, but complexity of managing and detecting failures worsens

Engineering Contradiction:

Improvesystem resilienceVSAvoidcomplexity of managing and detecting failures

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements feedback mechanisms where storage units send status information to a managing unit, which tracks the health of each storage unit. When failures are detected, the system automatically triggers recovery processes by sending requests to replace failed storage units and migrate data, creating a closed-loop feedback system that reduces manual management complexity.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent introduces a managing unit as an intermediary between storage units and the control system. This managing unit consolidates failure detection, analysis, and coordination functions, acting as a mediator that simplifies the complexity of managing dispersed storage units by centralizing monitoring and response coordination.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If error correction encoding is applied to enable failure recovery, then data integrity improves, but processing time and computational resources worsen

Engineering Contradiction:

Improvedata integrityVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent applies error correction encoding (Cauchy Reed-Solomon) to data before dispersing it across storage units. This preliminary encoding action enables automatic recovery from failures without requiring complex real-time processing during failure events, as the recovery process simply retrieves and decodes pre-encoded data slices.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent divides data into multiple encoded slices distributed across different storage units. This segmentation allows parallel processing during recovery operations, where multiple slices can be retrieved and reassembled simultaneously, reducing overall recovery time compared to processing entire data sets sequentially.

Inventive Principle:

Principle #1Segmentation

3Productivity

If multiple vaults share common storage units to improve resource utilization, then storage efficiency improves, but impact of failures worsens

Engineering Contradiction:

Improvestorage efficiencyVSAvoidimpact of failures

Core Design Contradiction:

ProductivityVSObject-affected harmful factors

Solution Approach 1:

The patent implements feedback mechanisms where the managing unit tracks which storage units are shared across multiple vaults and monitors their health status. When a failure occurs in a shared storage unit, the system automatically identifies all affected vaults and initiates coordinated recovery processes, providing feedback-driven response that mitigates the amplified impact of failures.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent employs error correction encoding with redundant data slices stored across multiple storage units. This beforehand cushioning ensures that even when shared storage units fail, the system has pre-prepared redundant copies and encoding schemes that can recover data, cushioning against the amplified impact that would otherwise affect multiple vaults.

Inventive Principle:

Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentUS11650878B2Failure abatement approach for a failed storage unit

Publication Date: 2023.05.16 PURE STORAGE INC

AI summary

A method for execution by a vault management device of a storage network includes determining a failure impact level to vaults of the storage network based on a failed storage unit within the vaults, where the vaults include a first vault that is associated with a first set of storage units and a first decode threshold number, and a second vault that is associated with a second set of storage units and a second decode threshold number, and where the failure impact level is based on the number of non-failed storage units within each of the vaults. The method continues with determining a failure abatement approach based on the failure impact level. The method continues by with facilitating the failure abatement approach.