Exact Regeneration of Failed Nodes in Distributed Storage

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing distributed storage systems face challenges in exactly regenerating a failed storage node with minimal computational complexity, especially when only a limited number of nodes are accessible, and struggle to achieve systematic repair for redundant nodes.

Innovation Solution

The method involves splitting an original file into segments, encoding them using XOR operations, and storing these segments across nodes, allowing for efficient reconstruction and regeneration of failed nodes using a processor, encoder, and decoder, with a focus on reduced computational complexity and partial file recovery capabilities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If erasure coding techniques are used for data storage, then reliability is improved, but computational complexity increases

Engineering Contradiction:
Improvedata reliabilityVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent divides the original file into multiple file segments and stores them across different storage nodes. When a node fails, only the specific segments stored on that node need to be regenerated from the remaining nodes, rather than regenerating everything. This segmentation approach reduces the computational complexity of regeneration while maintaining reliability through distributed storage of segments.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts only the necessary information for regeneration by identifying which file segments are stored on failed nodes and which are available on active nodes. The regeneration process then extracts only these specific segments from active nodes to replace failed ones, avoiding the need to process the entire file and reducing computational overhead.

Inventive Principle:
Principle #2Taking out (Extraction)

2Ease of repair

If systematic repair is implemented for redundant nodes, then repair capability is improved, but exact regeneration becomes difficult to achieve

Engineering Contradiction:
Improverepair capabilityVSAvoidexact regeneration accuracy
Core Design Contradiction:
Ease of repairVSManufacturing precision

Solution Approach 1:

The patent uses copying to regenerate failed nodes by directly copying the file segments from active storage nodes to the failed node's replacement. This copying approach ensures exact regeneration of the failed node's data without requiring complex reconstruction operations, thereby achieving both ease of repair and exact regeneration accuracy.

Inventive Principle:
Principle #26Copying

3Reliability

If more storage nodes are used for redundancy, then reliability is improved, but communication and processing costs increase

Engineering Contradiction:
Improvesystem reliabilityVSAvoidcommunication and processing costs
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent applies partial action by regenerating only the necessary file segments for failed nodes rather than regenerating all segments across all nodes. This partial regeneration approach reduces communication bandwidth and processing requirements while maintaining the reliability benefits of having multiple storage nodes for redundancy.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS8775860B2System and method for exact regeneration of a failed node in a distributed storage system
Publication Date: 2014.07.08 TATA CONSULTANCY SERVICES LTD
  • US8775860B2 patent drawing
  • US8775860B2 patent drawing
  • US8775860B2 patent drawing

AI summary

The invention relates to a method and a system for regenerating a failed storage node from one or more storage nodes storing an original file and also reconstructing the said original file. The method involves encoding the file segments using a XOR operation, decoding the encoded file segment using the XOR operation and regenerating the failed node by using the storage nodes which are in active position. The regenerated file segments are further written to a new storage node.