Storage Controller RAID Conversion to Distributed Parity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional RAID configurations using parity face difficulties in migrating to distributed RAID configurations, which can shorten rebuild times, due to the lack of easy conversion methods from parity-based to distributed RAID setups.
Innovation Solution
A storage system with a controller that generates parity data and allocates it across multiple drives, allowing for the conversion of conventional RAID configurations to distributed RAID by adding storage drives and redistributing data, ensuring data integrity and reducing rebuild times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If conventional parity-based RAID configuration is used, then storage capacity is efficiently utilized, but rebuild time is extended when a drive fails
Solution Approach 1:
The patent segments the rebuild process by introducing distributed RAID groups that distribute data and parity across multiple drives. When a drive fails, only a portion of the data needs to be rebuilt from distributed parity information across multiple drives, rather than rebuilding from a single parity drive as in conventional RAID, thereby reducing rebuild time
Solution Approach 2:
The patent transitions from conventional two-dimensional RAID architecture (data drives and parity drives in fixed groups) to a distributed multi-dimensional architecture where data and parity are spread across multiple drives in a grid-like structure, enabling parallel rebuild operations and reducing the time required for data recovery
2Loss of time
If distributed RAID configuration is implemented, then rebuild time is shortened, but conversion from conventional RAID is difficult
Solution Approach 1:
The patent introduces dynamic conversion capability that allows the storage system to transition from conventional RAID configuration to distributed RAID configuration on-demand. The system can dynamically reorganize data and parity distribution across drives, converting between architectures without requiring complete system redesign or data migration to external systems
Solution Approach 2:
The patent creates a universal storage system that can operate in both conventional RAID mode and distributed RAID mode. The same hardware infrastructure supports multiple RAID configurations, and the system can convert between them, providing multi-functionality that eliminates the need for separate systems for different RAID types
3Adaptability or versatility
If storage drives are added to convert from conventional RAID to distributed RAID, then data distribution is improved, but device complexity increases
Solution Approach 1:
The patent implements self-service automation where the storage controller automatically performs the complex tasks of data redistribution, parity recalculation, and drive reconfiguration when converting from conventional to distributed RAID. The system manages its own transformation process without requiring manual intervention, reducing the perceived complexity for users
Data Source
AI summary
A storage system including a storage controller and a plurality of storage drives generates parity data from data. Data and parity data configure a stripe. A plurality of stripes configure a parity group allocated with a plurality of storage drives to store the parity group. A first parity group allows the number of storage drives allocated to the parity group to be equal to the number of data pieces configuring each stripe. A second parity group allows the number of storage drives allocated to the parity group to be larger than the number of data pieces configuring each stripe and allows data for each stripe to be distributed and stored in different combinations of storage drives. When the first parity group is converted to the second parity group, a storage drive is added so that it is allocated to the parity group. Data is moved to ensure a free area.