Distributed Storage Parity Fragmentation for Disaster Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data storage systems for disaster recovery require excessive storage space, often doubling the original data size for redundancy, which is inefficient and costly, especially when protecting against site outages and data loss across geographically dispersed locations.
Innovation Solution
A method and system that distribute data and parity information across three storage units at different sites, where data is divided into parts and parity data is computed and stored across these units, allowing for efficient recovery with a storage requirement of one and a half times the original data size, while maintaining fault tolerance and reducing storage needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is copied from a first site to a second site for disaster recovery, then data protection against site outages is improved, but storage space requirement doubles to twice the original data size
Solution Approach 1:
The patent divides data into multiple fragments (first data fragment, second data fragment, third data fragment) and distributes them across different storage units at different sites. Instead of storing complete copies of all data at each site, each storage unit holds only a portion of the fragmented data, significantly reducing the storage space required at each location while maintaining the ability to reconstruct the complete data set.
Solution Approach 2:
The patent transforms the storage architecture by changing the parameter of data representation from complete copies to fragmented portions with associated parity information. By storing data fragments combined with parity data (XOR-based redundancy) across distributed storage units, the system achieves the same fault tolerance as full replication but with reduced storage requirements of approximately 1.5 times the original data size.
2Ease of repair
If complete data copies are stored at multiple sites, then data recovery capability is improved, but storage cost and efficiency deteriorate
Solution Approach 1:
The patent segments data into multiple fragments distributed across different storage units. When data loss occurs at one site, the system can recover the lost data fragments by retrieving the remaining fragments from other sites and using the parity information to reconstruct the missing portions, eliminating the need to store complete data copies at every location.
Solution Approach 2:
The patent introduces parity data as an intermediary element that enables data recovery without requiring complete copies of the original data at each site. The parity data, generated through XOR operations on data fragments, acts as a mediator that allows reconstruction of lost data fragments when combined with the remaining fragments from distributed storage units.
3Quantity of substance
If data is distributed across three sites with parity information, then storage efficiency is improved to one and a half times original size, but system complexity increases
Solution Approach 1:
The patent divides data into three fragments (first, second, and third data fragments) and distributes them across three different storage units at different sites. This segmentation approach achieves storage efficiency of 1.5 times the original data size while managing system complexity through systematic distribution and recovery protocols.
Solution Approach 2:
The patent changes the storage parameter from complete data copies to a combination of data fragments and parity information. This parameter transformation enables the system to achieve improved storage efficiency while maintaining manageable complexity through the use of XOR-based parity calculations and structured data reconstruction protocols.
Data Source
AI summary
Provided is a method in which data is received at the first storage unit. A first information unit, a second information unit, and a third information unit are generated, wherein the first information unit, the second information unit, and the third information unit each include a portion of the received data and computed parity data. The first information unit is stored in the first storage unit at the first site. The second information unit is distributed to the second storage unit at the second site for storage. The third information unit is distributed to the third storage unit at the third site for storage.