Exact-Repair Regenerating Codes With Lower Repair Bandwidth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed storage systems face inefficiencies in data recovery and node repair due to high repair-bandwidth requirements and complex combinatorial constraints, particularly in achieving the optimal trade-off between per-node storage and repair-bandwidth for exact-repair regenerating codes, especially for general system parameters (n, k, d).
Innovation Solution
A novel coding scheme for exact-repair regenerating codes that offers a unified storage-vs.-bandwidth trade-off, allowing for flexible system expansion and independent helper participation in the repair process, characterized by a determinant code construction that achieves optimal storage and repair bandwidth without depending on the total number of nodes, and reduces repair bandwidth for multiple failures through sequential repair mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional erasure codes are used for data redundancy, then fault tolerance is achieved, but repair-bandwidth becomes excessively high due to the total size of data to be downloaded for repair of each failed node
Solution Approach 1:
The patent segments the repair process by having each helper node transmit only a specific subset of encoded symbols (repair symbols) rather than the entire node content. This segmentation is achieved through the deterministic code construction where each helper node computes and transmits only the necessary portion of data required for repair, significantly reducing the total repair-bandwidth while maintaining fault tolerance.
Solution Approach 2:
The patent extracts and transmits only the essential repair symbols from each helper node instead of the complete node content. The extraction process identifies and transmits only those specific encoded symbols that are necessary for reconstructing the failed node's content, thereby reducing unnecessary data transmission and optimizing repair-bandwidth efficiency.
2Productivity
If regenerating codes are used to reduce repair-bandwidth, then repair efficiency improves, but achieving optimal storage-vs.-bandwidth trade-off becomes complex due to combinatorial constraints for general system parameters (n, k, d)
Solution Approach 1:
The patent changes the fundamental parameters of the coding scheme by using deterministic codes with specific algebraic structures (e.g., based on finite fields and polynomial representations). This parameter change allows the system to achieve optimal storage-vs.-bandwidth trade-off for general (n, k, d) parameters without being constrained by complex combinatorial optimizations, as the algebraic structure provides direct formulas for code construction and repair operations.
Solution Approach 2:
The patent creates a universal coding scheme that works for general system parameters (n, k, d) without requiring parameter-specific constructions. The deterministic code framework provides a unified approach that handles various system configurations (different numbers of total nodes, minimum recovery nodes, and helper nodes) through the same algebraic principles, eliminating the need for complex case-by-case combinatorial designs.
3Reliability
If replication is used for redundancy, then system reliability improves, but storage efficiency deteriorates due to the high cost of extra storage units required
Solution Approach 1:
The patent uses composite coding structures that combine multiple encoded symbols from different nodes to form the redundant information. Instead of simple duplication (replication), the system creates composite encoded representations where each node stores a unique combination of encoded symbols. This composite approach provides the same fault tolerance as replication but with significantly reduced storage overhead, as each node contributes a fraction of the total data rather than a complete copy.
Data Source
AI summary
A distributed storage system includes a plurality of nodes comprising a first node, wherein a total number of nodes in the distributed storage system is represented by n, wherein a file stored in the distributed storage system is recovered from a subset of a number of nodes represented by k upon a file failure on a node in the distributed storage system, and wherein a failed node in the plurality of nodes is recovered from a number of helper nodes of the plurality of nodes represented by d. Upon detecting a failure in the first node, each helper node of the number of helper nodes is configured to determine a repair-encoder matrix, multiply a content matrix by the repair-encoder matrix to obtain a repair matrix, extract each linearly independent column of the repair matrix, and send the linearly independent columns of the repair matrix to the first node.