Distributed Storage With Multi-Reliability Erasure Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed storage systems face inefficiencies in data recovery and repair due to single-level reliability, which does not account for heterogeneous reliability requirements, leading to suboptimal storage and repair-bandwidth tradeoffs.
Innovation Solution
The implementation of multi-reliability regenerating (MRR) codes, which mix symbolic coefficients across data messages to achieve improved reliability and efficiency in data reconstruction and repair, using encoding and decoding functions to ensure data integrity across multiple storage nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If single-level reliability distributed storage is used, then system simplicity is maintained, but data recovery efficiency and reliability are suboptimal
Solution Approach 1:
The patent segments the distributed storage system into multiple reliability levels (first level, second level, etc.), where each level has different reliability parameters. Data is stored across these hierarchical levels, allowing the system to achieve high reliability for critical data while maintaining lower complexity for less critical data, thus resolving the contradiction between reliability and complexity.
Solution Approach 2:
The patent applies local quality by assigning different reliability characteristics to different portions of the storage system. Each storage node or data block can be assigned to different reliability levels based on its importance, allowing localized optimization of reliability without requiring the entire system to operate at maximum complexity and cost.
2Reliability
If higher reliability is achieved through traditional distributed storage, then data safety improves, but storage bandwidth and repair bandwidth increase suboptimally
Solution Approach 1:
The patent implements dynamic reliability adjustment where the system can adaptively select which reliability level to use based on current operational conditions, data importance, and resource availability. This dynamic approach allows the system to achieve necessary data safety while minimizing storage and repair bandwidth consumption by not always operating at maximum reliability levels.
Solution Approach 2:
The patent changes the reliability parameter by introducing multiple reliability levels with different characteristics. By selecting appropriate reliability levels for different data types and operational contexts, the system achieves data safety requirements while optimizing storage and repair bandwidth usage, avoiding the inefficiency of uniformly high reliability across all data.
3Stability of the object's composition
If uniform reliability is applied across all data, then system consistency is maintained, but hardware costs and storage efficiency deteriorate
Solution Approach 1:
The patent applies local quality by allowing different reliability levels for different data portions while maintaining overall system consistency through standardized interfaces and protocols. This enables the system to use appropriate hardware resources for each data type's actual requirements, avoiding waste of hardware resources on data that doesn't require high reliability while maintaining system-wide consistency.
Solution Approach 2:
The patent creates a universal multi-level reliability framework that can handle different reliability requirements within a single system architecture. This multi-functional approach allows the same storage infrastructure to serve both high-reliability and lower-reliability data needs, optimizing hardware resource utilization while maintaining system consistency through unified management.
Data Source
AI summary
Multi-reliability regenerating (MRR) erasure codes are disclosed. The erasure codes can be used to encode and regenerate data. In particular, the regenerating erasure codes can be used to encode data included in at least one of two or more data messages to satisfy respective reliability requirements for the data. Encoded portions of data from one data message can be mixed with encoded or unencoded portions of data from a second data message and stored at a distributed storage system. This approach can be used to improve efficiency and performance of data storage and recovery in the event of failures of one or more nodes of a distributed storage system.