Multi-dimensional Storage Pool for RAID Data Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current RAID-based data storage systems lack the ability to individually address storage units across different redundancy groups within a higher storage system tier, limiting data recovery speed and increasing the likelihood of data loss upon failure.
Innovation Solution
Implementing a multi-dimensional allocation structure where storage units are assigned to multiple redundancy groups, allowing for individual addressing and forming groups in a two- or three-dimensional structure, enabling faster data restoration and reduced system failure probability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If storage units are organized in traditional RAID groups with hierarchical structure, then redundancy and data protection are provided, but storage units cannot be individually addressed across different redundancy groups and data recovery speed is limited
Solution Approach 1:
The patent transitions from traditional hierarchical RAID groups to a multi-dimensional storage pool structure where storage units can be individually addressed across multiple dimensions. This allows storage units to participate in multiple redundancy groups simultaneously, enabling faster data recovery by accessing redundant data from any available storage unit in the pool rather than being constrained to specific hierarchical levels.
Solution Approach 2:
Storage units in the patent are designed to be universally addressable and can serve multiple functions across different redundancy groups. Each storage unit can be part of multiple redundancy groups simultaneously, allowing it to provide redundancy services to multiple data sets and enabling more flexible and faster data recovery operations.
2Reliability
If storage units are assigned to multiple redundancy groups, then data recovery speed and system reliability are improved, but system complexity increases
Solution Approach 1:
The patent segments the storage system into a pool of individually addressable storage units that can be dynamically assigned to multiple redundancy groups. This segmentation allows the system to manage complexity by treating each storage unit as an independent, addressable entity rather than being bound to rigid hierarchical structures, simplifying the management of multi-dimensional redundancy assignments.
3Reliability
If traditional hierarchical RAID groups are used, then data protection is provided, but the likelihood of data loss increases upon failure due to limited recovery options
Solution Approach 1:
The patent introduces multi-dimensional addressing and assignment of storage units to multiple redundancy groups, creating additional dimensions for data recovery. When a storage unit fails, the system can retrieve redundant data from any available storage unit across multiple redundancy groups, significantly increasing recovery flexibility and reducing data loss risk compared to traditional hierarchical structures.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The system has multiple storage units (1.1-1.r, 2.1-r.1) e.g. flash memory and hard disk, attached to storage system planes. The storage units are individually addressed by a storage system management part in the storage system planes. Linear redundant groups (10-50) and column redundant groups (100-500) of storage units are formed on the system planes. One of the storage units is attached to one of the redundant groups. Two of the redundant groups are connected to the storage unit, where the redundant groups are formed as a redundant array of independent disks group and a parity group. Independent claims are also included for the following: (1) a method for operating a data storage system (2) a computer program with a set of instructions for performing the method for operating the data storage system (3) a computer program product with a set of instructions for performing the method for operating the data storage system.