RAID Controller Metadata Recovery with Preassigned Strip Ownership
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current RAID storage systems experience increased temporary service suspension times due to the replication process when a controller fails, which is necessary to establish a new synchronization relationship, affecting space utilization and synchronization speed.
Innovation Solution
A metadata recovery method involving setting sorting and backup controllers, selecting and numbering strips, and establishing unique owner controllers to facilitate efficient metadata backup and recovery, reducing the need for recalculating ownership relationships during failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data replication is performed to establish a new synchronization relationship when a controller fails, then the RAID system maintains high availability, but the temporary service suspension time increases
Solution Approach 1:
The patent assigns unique owner controllers to specific strips in advance and pre-establishes backup controller relationships before any failure occurs. When a controller fails, the system can immediately activate the pre-designated backup controller without needing to perform dynamic replication or recalculate ownership relationships, thus eliminating service suspension time while maintaining high availability
2Reliability
If cycle mirroring is used to backup metadata across four controllers, then fault tolerance is improved, but space utilization rate decreases
Solution Approach 1:
The patent implements selective metadata backup where each controller only stores metadata for its assigned owner strips rather than duplicating all metadata across all controllers. This localized approach maintains fault tolerance through targeted backup while significantly improving space utilization by eliminating redundant metadata storage
3Adaptability or versatility
If dynamic recalculation of ownership relationships is performed during controller failure, then adaptability is maintained, but synchronization speed decreases
Solution Approach 1:
The system pre-assigns unique owner controllers to each strip and establishes backup relationships before failure occurs. This eliminates the need for dynamic recalculation during failure events, allowing immediate activation of backup controllers and maintaining high synchronization speed while preserving adaptability through pre-configured failover paths
Data Source
AI summary
Provided are a metadata recovery method and apparatus for a Redundant Array of Independent Disks (RAID) controller, a device, and a nonvolatile readable storage medium. The method includes: setting sorting of each controller in the RAID controller, and setting a backup controller for each controller according to the sorting; selecting a preset number of strips from each controller, numbering each strip, and setting a unique owner controller for numbered strips of each controller; backing up other controllers according to the set backup controller and the unique owner controller for the strips in response to a controller failure; and recovering a controller according to the set backup controller and the unique owner controller for the strips in response to controller recovery.