Distributed Storage Nodes Using Local Redundant Code Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed storage systems, data retrieval from multiple computer nodes leads to increased delay times and decreased throughput due to network transmission requirements, and recovery processes after node failures also face similar performance issues.
Innovation Solution
The system organizes data into user data and redundant code regions within logical storage areas on each computer node, generating group information to optimize data transmission and recovery processes, allowing for efficient write and read operations, and dynamically adjusts configurations to maintain redundancy without network reliance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is distributed and stored across multiple computer nodes, then data reliability is improved through redundancy, but network transmission delay increases and throughput deteriorates
Solution Approach 1:
The patent segments data into user data and redundant code, organizing them into distinct regions (user data region and redundant code region) within logical storage on each computer node. This segmentation allows the system to maintain distributed storage for reliability while enabling local access to both data and its redundant code, reducing network transmission delays.
2Reliability
If data is recovered by distributing it across multiple computer nodes, then data recovery capability is improved, but network transmission requirements increase and performance decreases
Solution Approach 1:
The patent merges the user data and redundant code into a unified logical storage structure on each computer node. By combining both data and its redundant code locally, the system enables recovery operations to be performed using local resources without requiring network transmission, thus improving data recovery performance while maintaining recovery capability.
3Reliability
If redundant code is stored on separate computer nodes, then data protection is improved, but system complexity increases due to configuration management
Solution Approach 1:
The patent makes each computer node universal by enabling every node to store both user data and redundant code in its local logical storage. This multi-functionality eliminates the need for specialized roles (data nodes vs. redundancy nodes), simplifying configuration management while maintaining robust data protection through distributed redundancy.
Data Source
AI summary
The performance of a computer system which distributes and stores user data and a redundant code is improved. Each computer generates group information indicating positions of a user data region and a redundant code region in each of a plurality of computers; when a write request for write data is received, each computer writes the write data to a local storage device, selects a transmission destination computer on the basis of the group information, and transmits transmission data to the transmission destination computer; when a plurality of pieces of transmission data are respectively received, each computer generates a redundant code by using the plurality of pieces of transmission data on the basis of the group information, and writes the redundant code to the local storage device; and when configurations of the plurality of computers are changed, each computer changes the group information on the basis of the changed configurations.