Distributed Storage Data Slicing for Fault Tolerance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current distributed storage systems face challenges in ensuring data integrity and availability across multiple storage units, particularly in handling large datasets and complex tasks, as they lack efficient error correction and task processing mechanisms.

Innovation Solution

A distributed computing system that employs dispersed error encoding and decoding, allowing data to be segmented, encoded, and distributed across multiple geographically disparate storage units, enabling robust storage and task processing with error correction and secure retrieval, while also facilitating secure and scalable data storage and task execution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data is stored across multiple dispersed storage units, then data availability and fault tolerance are improved, but system complexity and error correction requirements increase

Engineering Contradiction:
Improvedata availabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments data into multiple data slices and distributes them across different storage units. The data is divided into a first set of data slices stored at first storage units and a second set of data slices stored at second storage units, enabling fault tolerance while maintaining manageable complexity through systematic segmentation

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements preliminary error correction by encoding data before storage. Error correction codes are applied to data slices before they are distributed to storage units, allowing the system to tolerate failures without requiring complex real-time correction mechanisms, thus improving reliability while controlling complexity

Inventive Principle:
Principle #10Preliminary action

2Reliability

If error correction codes are applied to dispersed storage, then data integrity is improved, but processing time and computational overhead increase

Engineering Contradiction:
Improvedata integrityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the error correction process by applying codes to individual data slices rather than entire datasets. This segmented approach allows parallel processing and reduces computational overhead while maintaining data integrity across all slices

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent optimizes error correction parameters including the number of data slices, code rates, and distribution patterns to balance integrity requirements with processing efficiency. By adjusting these parameters, the system achieves reliable data integrity without excessive computational overhead

Inventive Principle:
Principle #35Parameter changes

3Productivity

If large datasets are processed in distributed manner, then task processing capability is improved, but coordination complexity and communication overhead increase

Engineering Contradiction:
Improvetask processing capabilityVSAvoidcoordination complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent segments large datasets into manageable data slices that can be independently processed across distributed storage units. This segmentation enables parallel task execution and improves processing capability while reducing coordination complexity through localized operations

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates multiple copies of data slices at different storage units, enabling redundant processing and improving task processing capability. The distributed copies allow independent task execution while reducing the need for complex coordination through replication

Inventive Principle:
Principle #26Copying

Data Source

PatentUS20240430323A1Balancing throughput and response time quality of service
Publication Date: 2024.12.26 PURE STORAGE INC
  • US20240430323A1 patent drawing
  • US20240430323A1 patent drawing
  • US20240430323A1 patent drawing

AI summary

A method for execution by one or more computing devices includes scheduling processing of one or more data access requests of a plurality of data access requests to produce an execution schedule, where the scheduling processing is in accordance with an access performance level for previous data access requests and in accordance with a desired access performance level. The method further includes executing the one or more data access requests in accordance with the execution schedule. The method further includes determining an updated access performance level compares unfavorably to the desired access performance level. The method further includes implementing an alternate throughput scheme for subsequent data access requests of the plurality of data access requests.