RAID Controller Logical Unit Segmentation for Drive Rebuild
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional RAID apparatuses face challenges in reducing the time and cost associated with rebuilding due to the need for large-capacity management tables and extensive processing times, even though they allow continued use of HDDs by inhibiting replacement.
Innovation Solution
A drive array apparatus with a RAID controller that divides data storage areas into logical units, manages these units, and performs reconfiguration by identifying error occurrences based on logical unit areas, reducing the need for HDD replacement and minimizing table information requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a conventional management table is used to manage all LBA corresponding to each head, then rebuilding allows continued use of HDDs, but the management table requires large capacity and processing time increases
Solution Approach 1:
The patent segments the management table into a RAID configuration management table (stored in the controller) and a LUN/LBA table (stored in each HDD). The RAID configuration management table only stores necessary mapping relationships between logical unit numbers and physical disk locations, while individual HDDs store their own LUN/LBA mappings. This segmentation dramatically reduces the size of the centralized management table while maintaining the ability to continue using HDDs during rebuilding.
Solution Approach 2:
The patent extracts the LUN/LBA table from the centralized management table and stores it within each HDD individually. This extraction removes the burden of storing complete LBA mappings for all heads in the centralized table, reducing its capacity requirements while preserving the functionality to track and manage individual disk locations during the rebuilding process.
2Ease of operation
If a conventional management table is used to manage all LBA corresponding to each head, then rebuilding allows continued use of HDDs, but processing time to reference and update the management table increases
Solution Approach 1:
By segmenting the management table into a centralized RAID configuration management table and individual LUN/LBA tables stored in each HDD, the patent reduces the amount of data that needs to be referenced and updated in the centralized table during rebuilding operations. The centralized table only needs to maintain high-level mapping relationships, significantly reducing processing time while enabling continued HDD usage.
Solution Approach 2:
The patent performs preliminary action by having each HDD store its own LUN/LBA table locally before rebuilding operations begin. This pre-stored information eliminates the need to search through a large centralized management table during rebuilding, as the necessary LBA mappings are already available in the failing HDD's local table, thereby reducing processing time.
3Reliability
If HDD replacement is performed, then storage reliability is maintained, but execution time and operational burden increase
Solution Approach 1:
The patent applies the principle of discarding and recovering by enabling the continued use of failing HDDs during the rebuilding process. Instead of immediately replacing and discarding the failing drive, the system recovers and utilizes its data by copying it to other drives while the failing HDD remains online and accessible, thereby maintaining reliability while reducing the time and operational burden associated with complete drive replacement.
Data Source
AI summary
According to one embodiment, a drive array apparatus includes a plurality of data storage drives constituting a drive array and a controller. The controller divides all data storage areas of each of the data storage drives into a plurality of logical unit areas and associates and manages each of the logical unit areas and each of the data storage drives. The controller identifies a logical unit area of an error occurrence object based on a notification from a data storage drive where an error occurred, restores data of the identified logical unit area, and performs reconfiguration processing of the drive array.