Balanced Data Segment Distribution for Storage Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage systems using RAID technology face challenges in restoring data when multiple segments are lost due to disk failures, as data distribution is not always evenly balanced across storage nodes, leading to incomplete recovery.
Innovation Solution
A method and apparatus for data storage, distribution, and recovery that assigns storage objects based on a predefined erasure coding strategy (k+m), ensuring that each storage object holds a balanced number of data segments, with the difference between objects being within a preset threshold, allowing for even distribution and restoration in case of failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data segments are randomly stored in disks, then storage flexibility is improved, but data recovery reliability deteriorates when multiple segments are lost
Solution Approach 1:
The patent divides data into multiple segments and distributes them across different storage objects in a balanced manner. Each storage object contains a controlled number of segments, ensuring that no single storage object holds too many segments. This segmentation approach allows the system to maintain flexibility in storage allocation while ensuring that data can be recovered even if some storage objects fail, as long as the number of failed objects does not exceed the redundancy threshold.
Solution Approach 2:
The patent implements local quality control by ensuring that each storage object has a similar number of data segments stored in it. This is achieved by distributing segments uniformly across storage objects, with the difference between any two storage objects being at most one segment. This uniform distribution ensures that no single storage object becomes a critical point of failure, thereby improving overall data recovery reliability while maintaining storage flexibility.
2Productivity
If multiple data segments are stored in one disk, then storage space utilization is improved, but data loss risk increases when the disk fails
Solution Approach 1:
The patent segments data into multiple smaller units and distributes them across multiple storage objects rather than concentrating them in a single disk. This segmentation ensures that even if one disk fails, only a limited number of segments are lost, and the remaining segments can be used to reconstruct the original data through erasure coding, thereby reducing data loss risk while maintaining efficient storage space utilization.
Solution Approach 2:
The patent changes the parameter of data distribution by implementing a balanced distribution strategy where each storage object contains a controlled number of segments. This parameter change from concentrated storage to balanced distribution reduces the risk of data loss while maintaining storage efficiency, as the system can tolerate failures of up to m storage objects out of k+m total objects.
3Reliability
If data is equally distributed across storage objects, then data recovery capability is improved, but storage management complexity increases
Solution Approach 1:
The patent segments both data and storage resources into uniform units that can be easily managed. By dividing data into segments and storage objects into uniform containers, the system achieves balanced distribution through simple rules rather than complex algorithms. This segmentation approach improves data recovery capability while keeping storage management relatively simple through standardized operations.
Solution Approach 2:
The patent implements homogeneity in the storage system by ensuring all storage objects have the same structure and capacity to hold segments. This homogeneous design simplifies storage management operations such as allocation, retrieval, and failure handling, while simultaneously improving data recovery capability through uniform distribution. The homogeneous structure allows for straightforward implementation of the balanced distribution strategy without requiring complex management mechanisms.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
Embodiments of the present application provide a method and apparatus for displaying objects. In the method, data segments of to-be-stored data are stored in storage objects. A storage object is a minimum unit for data storage. The difference between the number of data segments stored in a storage object and that of another storage object is no more than a first preset threshold. Data segments are equally stored, to the greatest extent, into storage objects. The case that data having segments lost due to failure of a storage object cannot be restored is avoided.