Predictable RAID Member Assignment for Flexible Storage Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage systems face inefficiencies in scaling storage capacity by adding new disks, as the increment of new disks required for RAID configurations becomes large and inefficient with technological advancements.
Innovation Solution
A storage array with a disk manager that creates and grows disk clusters by distributing RAID protection group members in predictable patterns across multiple disks, allowing for efficient addition of new disks and validation of disk clusters, using modes that match member numbers with disk ordinals or split indices, and designating spare splits for failure recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If storage capacity is increased by adding W new disks for RAID (D+P) implementation, then storage capacity is improved, but the increment of new disks becomes large and inefficient
Solution Approach 1:
The patent segments each physical disk into multiple logical splits (e.g., 4 splits per disk), transforming the scaling unit from entire disks to fractional splits. This allows the RAID array to scale in smaller, more efficient increments by utilizing portions of existing disks rather than requiring complete disk additions.
Solution Approach 2:
The patent introduces a new dimension of organization by distributing RAID protection group members across split indices rather than solely across disk ordinals. This multidimensional distribution approach (disk ordinal × split index) enables more flexible and granular scaling compared to traditional disk-based expansion.
2Ease of manufacture
If individual disks are used as RAID protection group members, then simplicity of implementation is improved, but flexibility in scaling is worsened
Solution Approach 1:
By dividing each disk into multiple splits, the system maintains the simplicity of disk-based management while gaining the flexibility of finer-grained allocation. Each split can be independently assigned to RAID protection groups, enabling adaptable scaling without complex cross-disk coordination.
Solution Approach 2:
The patent changes the fundamental parameter of allocation from whole-disks to fractional-splits, transforming the system from rigid disk-based scaling to flexible split-based scaling while maintaining compatibility with existing RAID methodologies.
3Ease of operation
If protection group members are distributed without predictable patterns, then ease of initial configuration is improved, but validation and error detection become difficult
Solution Approach 1:
The patent establishes predictable distribution patterns (mode 1 and mode 2) as preliminary rules before data is written to the RAID array. These predetermined patterns enable automatic validation of disk clusters by checking whether member assignments conform to the expected mathematical relationships between disk ordinals and split indices.
Solution Approach 2:
The system incorporates validation mechanisms that provide feedback on whether protection group member assignments conform to the predictable patterns. This feedback loop ensures data integrity by detecting configuration errors or corruption through mathematical verification of member distribution.
Data Source
AI summary
A flexible RAID system can expand by iteratively adding disks to a disk cluster. The disks may be added individually or in groups. Spare capacity is distributed over multiple disks within a cluster. Clusters are created, grown, and divided with predictable RAID member distribution patterns. Cluster validation is performed based on the patterns.