Multi-Site Erasure Coding for Storage-Efficient Data Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network-based backup systems require exceptionally reliable centralized backup sites, leading to inefficient storage needs as duplicating data for redundancy doubles the required storage space.
Innovation Solution
Implementing erasure encoding, such as Reed/Solomon encoding, to distribute data across multiple sites, allowing for error detection and correction without duplicating storage space, where data is encoded into components and stored across separate sites, enabling efficient data redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is duplicated for redundancy at the centralized backup site, then reliability is improved, but storage space requirements double
Solution Approach 1:
The patent segments the backup data into multiple components and distributes them across multiple backup sites. Instead of duplicating all data at one centralized site, the data is divided and stored across several sites, reducing the storage burden on any single site while maintaining redundancy through distribution.
Solution Approach 2:
The patent implements nested encoding by applying erasure encoding to data components stored at multiple sites. Each site stores encoded components that contain redundant information, allowing recovery of original data if some sites fail. This nested redundancy structure provides reliability without requiring full duplication at each level.
2Quantity of substance
If data is distributed across multiple sites using erasure encoding, then storage efficiency is improved, but system complexity increases
Solution Approach 1:
The patent uses erasure encoding to create redundant copies of data components that are distributed across multiple sites. Rather than simple duplication, the encoding creates mathematically derived copies that can reconstruct the original data, providing storage efficiency through compact representation while managing complexity through standardized encoding algorithms.
3Ease of operation
If a centralized backup site is used, then ease of operation is improved, but reliability deteriorates due to single point of failure
Solution Approach 1:
The patent segments the centralized backup architecture into multiple distributed backup sites. This segmentation eliminates the single point of failure while maintaining ease of operation through automated encoding and distribution processes that manage the complexity of multiple sites.
Solution Approach 2:
The patent introduces an intermediary encoding process that mediates between the client and multiple backup sites. This intermediary layer handles the complexity of distribution and redundancy, allowing clients to operate simply while the system manages the distributed architecture behind the scenes.
Data Source
AI summary
Storing data includes encoding the data using an erasure encoding to provide encoded data, separating the encoded data into a number of components, and providing each of the components to separate data sites. The data may be encoded using a Reed/Solomon encoding. The data may be encoded using a 3-1 Reed/Solomon encoding and the encoded data may be separated into four components that are provided to four separate sites. The data may be encoded by a site that receives the data prior to encoding. The data may be encoded by a client that provides the data to a plurality of sites coupled thereto. Storing data may also include each of the separate data sites providing an additional encoding for the components. The data and the components may be encoded using a Reed/Solomon encoding.