Erasure-Coded Data Storage With Flexible Redundancy and Fast Rebuilds

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing data storage methods for video surveillance, such as RAID and backup schemes, face inefficiencies in disk space usage, redundancy, and flexibility, particularly when handling data loss or corruption, and require significant additional capacity or fixed disk configurations.

Innovation Solution

A computer-implemented method that splits data into N pieces and generates M redundancy pieces, storing them on separate storage media at the same offset within a file set, allowing for flexible redundancy levels and efficient use of disk space, with the option to vary the number of disks and redundancy based on data importance and retention schedules.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If full duplicates of the data are stored (RAID1), then data reliability is improved, but disk capacity requirement doubles

Engineering Contradiction:
Improvedata reliabilityVSAvoiddisk capacity requirement
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The data is segmented into N pieces and distributed across multiple storage media, with only M redundancy pieces stored separately. This segmentation allows the system to achieve reliable data recovery without storing complete duplicates, reducing the total storage capacity requirement compared to RAID1 mirroring.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system changes the redundancy parameter from full duplication (RAID1) to erasure coding with configurable N and M values. By adjusting these parameters, the system can optimize the balance between data reliability and storage capacity utilization, achieving reliable data protection with lower storage overhead.

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If RAID 5 or 6 is used, then storage overhead is reduced, but flexibility in varying redundancy levels is lost

Engineering Contradiction:
Improvestorage overheadVSAvoidflexibility in redundancy levels
Core Design Contradiction:
Quantity of substanceVSAdaptability or versatility

Solution Approach 1:

The system implements dynamic redundancy configuration where N and M can be adjusted based on data importance and retention schedules. This dynamic approach allows the system to adapt redundancy levels for different data sets, providing flexibility that fixed RAID configurations cannot offer while maintaining efficient storage overhead through erasure coding.

Inventive Principle:
Principle #15Dynamics

3Device complexity

If fixed size disk sets are used, then system complexity is reduced, but ability to handle drive failures with varied disk availability is limited

Engineering Contradiction:
Improvesystem complexityVSAvoidhandling varied disk availability
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The erasure coding system with separate file sets and offsets provides a universal storage framework that can handle various disk availability scenarios. The same basic structure supports different N and M configurations, allowing the system to adapt to different numbers of available disks and failure scenarios without requiring complex specialized configurations for each case.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Reliability

If data is rebuilt after drive failure, then data reliability is restored, but performance bottlenecks occur

Engineering Contradiction:
Improvedata reliability restorationVSAvoiddata rebuild performance
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

By segmenting data into pieces stored at the same offset in separate files across multiple storage media, the system enables parallel reconstruction. When a drive fails, the redundant pieces from other drives can be read simultaneously and used to reconstruct the missing data, improving rebuild performance compared to sequential reconstruction methods.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3734457B1Method of storing items of data
Publication Date: 2023.08.23 VAION LTD

AI summary

A computer-implemented method of storing an item of data across a plurality of storage media, the method comprising the steps of: receiving an item of data to be stored; splitting the item of data into N pieces of data; generating M redundancy pieces of data, usable to rebuild the item of data; storing each of the N pieces of data and M redundancy pieces of data on separate storage media of the plurality of storage media at a same offset within a file of the respective storage medium, the files of the storage media containing the N pieces of data and the M redundancy pieces of data being associated as a file set; and storing, separately to the N pieces of data and M redundancy pieces of data, the offset and the file set.